Skip to main content

Full text of "Elektor Magazine"

See other formats


•VwVH ■•T' 
f— . I 


>—‘7 





advertisement 


elektor april 1978 — E 3 


E 

1977 ISSUES 




PRICE 
£2.30 
(incl. p&p) 


In 1977 we published over 200 different designs in Elektor which 
included approx. 29 Audio articles, 6 Car articles, 18 Design ideas 
18 Domestic circuits, 10 Fun, games and model building Circuits, 

18 Information articles microprocessors 

13 music circuits 12 Paranormal circuits 

13 Power supply circuits, 17 R.F. circuits, 37 Test and Measuring 
Equipment circuits, 7 Time, Timers and counters, circuits, and 
1 1 miscellaneous designs 

Due to our increased circulation and popularity of these articles, 
only limited amounts of 1977 back issues are available so please 

order your set of issues 21-32, while our stocks last 

we would hate to disappoint you 


A more detailed cumulative index for 1977 is published in the 
December 1977 issue number 32. A list of available printed circuit 
boards and their prices can be found in the EPS list at the front of 
this issue. 

To order these, please refer to 'Elektor readers service order form' 
in this issue. 


please don t 
leave your i 

issues of J 

just 
Mug 
about 


a professionally pro* 
duced magazine deser 
ves to be well kept, so 
keep your elektors in a 
binder, order your 1978 
binder now. available 
from elektor publishers 
ltd. 

These smart green binders will keep 
your copies of Elektor clean and tidy. 
Each issue is easily removed for reference. 



E-4 — elektor april 1978 


decoder 


elektor 36 decoder 

Volume 4 Number 4 


Editor 

Deputy editor 
Technical editors 


Subscriptions 


W. van der Horst 
P. Holmes 

J. Barendrecht, G.H.K. Dam, 

E. Krempelsauer, G.H. Nachbar 
A. Nachtmann, K.S.M. Walraven 
Mrs. A. van Meyel 


International head offices: Elektuur Publishers Ltd. 

Bourgognestr. 13a 
Beek (L), Netherlands 
Tel. 04402-4200 
Telex: 56617 Elekt NL 


U.K. editorial offices, administration and advertising: 

Elektor Publishers Ltd., Elektor House, 

10 Longport Street, Canterbury CT1 1PE, Kent. U.K. 

Tel.: Canterbury (0227)54430. Telex: 965504. 

Please make all cheques payable to Elektor Publishers Ltd. 
at the above address. 

Bank: 1. Midland Bank Ltd., Canterbury, A/C no. 11014587 
Sorting code 40-16-11, Giro no. 31 54254. 

2. U.S.A. only: Bank of America, c/o World Way 
Postal Center, P.O. Box 80689, Los Angeles, 

CA 90080, A/C no. 12350 04207. 

3. Canada only: The Royal Bank of Canada, 

c/o Lockbox 1969, Postal Station A, Toronto, 
Ontario, M5W 1W9. A/C no. 160-269-7. 

Assistant Manager and Advertising : R.G. Knapp 
Editorial : T. Emmens 


ELEKTOR IS PUBLISHED MONTHLY on the third Friday of each 
month. 

1. U.K. and all countries except the U.S.A. and Canada: 

Cover price £ 0.50. 

Number 39/40 (July/August), is a double issue, 

'Summer Circuits', price £ 1. — . 

Single copies (incl. back issues) are available by post from our 
Canterbury office, at £ 0.60 (surface mail) or £ 0.95 (air mail). 
Subscriptions for 1978, January to December incl., 

£ 6.75 (surface mail) or £ 12.00 (air mail). 

2. For the U.S.A. and Canada: 

Cover price $ 1 .50. 

Number 39/40 (July/August), is a double issue, 

'Summer Circuits', price S 3. — . 

Single copies (incl. back issues) S 1.50 (surface mail) or 
$ 2.25 (air mail). 

Subscriptions for 1978, January to December incl., 

S 18. — (surface mail) or S 27. — (air mail). 

All prices include post & packing. 

CHANGE OF ADDRESS. Please allow at least six weeks for change of 
address. Include your old address, enclosing, if possible, an address label 
from a recent issue. 

LETTERS SHOULD BE ADDRESSED TO the department concerned: 
TO = Technical Queries; ADV = Advertisements; SUB = Subscriptions, 
ADM = Administration; ED = Editorial (articles submitted for 
publication etc.); EPS = Elektor printed circuit board service. 

For technical queries, please enclose a stamped, addressed envelope or 
a self-addressed envelope plus an IRC. 

THE CIRCUITS PUBLISHED ARE FOR domestic use only. The sub- 
mission of designs or articles to Elektor implies permission to the 
publishers to alter and translate the text and design, and to use the 
contents in other Elektor publications and activities. The publishers 
cannot guarantee to return any material submitted to them. All 
drawings, photographs, printed circuit boards and articles published in 
Elektor are copyright and may not be reproduced or imitated in whole 
or part without prior written permission of the publishers. 

PATENT PROTECTION MAY EXIST in respect of circuits, devices, 
components etc. described in this magazine. 

The publishers do not accept responsibility for failing to identify such 
patent or other protection. 

National ADVERTISING RATES for the English-language edition of 
E ektor and/or international advertising rates for advertising at the same 
time in the English. Dutch and German issues are available on request. 
DISTRIBUTION in U.K.: Spotlight Magazine Distributors Ltd., 
Sootfight House 1, Bentwell Road, Holloway, London N7 7AX. 
DISTRIBUTION in CANADA: Gordon and Gotch (Can.) Ltd., 

S5 York Street, Toronto, Ontario, M5J 1S4. 

-opyright ©1978 Elektor publishers Ltd — Canterbury. 

; -ted in the Netherlands. 


What is a TUN? 

What is 10 n? 

What is the EPS service? 
What is the TQ service? 
What is a missing link? 


Semiconductor types 
Very often, a large number of 
equivalent semiconductors exist 
with different type numbers. For 
this reason, 'abbreviated' type 
numbers are used in Elektor 
wherever possible: 

• '741 ' stand for pA741 , 

LM741 , MC641 , MIC741, 
RM741, SN72741, etc. 

• 'TUP' or 'TUN’ (Transistor, 
Universal, PNPor NPN respect- 
ively) stand for any low fre- 
quency silicon transistor that 
meets the following specifi- 
cations: 


! UCEO, max 

20V 

• C, max 

100 mA i 

hfe, min 

100 

Ptot, max 

100 mW 

; fT, min 

100 MHz 

« < 

1 


Some 'TUN's are: BC107, BC108 
and BC109 families; 2N3856A 
2N3859, 2 N 3860, 2N3904, 
2N3947, 2N4124. Some 'TUP's 
are: BC1 77 and BC1 78 families; 
BC179 family with the possible 
exeption of BC1 59 and BC1 79; 
2N2412, 2N3251 . 2N3906, 
2N4126, 2N4291. 

• 'DUS' or 'DUG' (Diode Univer- 
sal, Silicon or Germanium 
respectively) stands for any 
diode that meets the following 
specifications: 



DUS 

Txjg 

UR, max 

25V 

[20V 

IF, max 

100mA 

35mA 

1 R, max 

1 pA 

100 pA 

Ptot, max 

250mW 

250mW 

Cp, max 

5pF 

[l OpF 


Some 'DUS's are: BA127, BA217, 
BA218, BA221 , BA222, BA317, 
BA318, BAX13, BAY61 , 1N914, 
1N4148. 

Some 'DUG's are: OA85, OA91 , 
OA95, AA116. 

• 'BC107B', 'BC237B', 'BC547B' 
all refer to the same 'family' of 
almost identical better-quality 
silicon transistors. In general, 
any other member of the same 
family can be used instead. 

BC107 (-8, -9) families: 

BC107 (-8. -9), BC147 (-8, -9) 
BC207 (-8, -9), BC237 (-8, -9) 
BC317 (-8. 9). BC347 (-8, -9) 
BC547 (-8, 9), BC171 (-2 -3)’ 
BC182 (-3, -4). BC382 (-3 -4) 
BC437 (-8, -9). BC414 

BC177 (-8, -9) families: 

BC177 (-8, -9), BC157 (-8, -9), 
BC204 (-5, -6), BC307 (-8, -9), 
BC320 (-1, -2), BC350 (-1, -2). 
BC557 (-8. -9), BC251 (-2, -3). 
BC212 (-3. 4). BC512 (-3. 4) 
BC261 (-2. -3). BC416. 

Resistor and capacitor values 
When giving component values, 
decimal points and large numbers 


of zeros are avoided wherever 
possible. The decimal point is 
usually replaced by one of the 
following abbreviations: 


P 

(pico-l = 

10 

n 

(nano-) = 

10 

M 

(micro ) - 

10 

m 

(milli-l = 

10 

k 

(kilo-) = 

10 

M 

(mega-) = 

10' 

G 

(giga) = 

10' 


A few examples: 

Resistance value 2k7: 2700 fl. 
Resistance value 470: 470 n. 
Capacitance value 4p7: 4.7 pF, or 
0.000 000 000 004 7 F . . . 
Capacitance value lOn: this is the 
international way of writing 

1 0.000 pF or .01 pF, since 1 n is 
1 0‘ ’ farads or 1 000 pF . 

Resistors are % Watt 5% carbon 
types, unless otherwise specified 
The DC working voltage of 
capacitors (other than electro- 
lytics) is normally assumed to be 
at least 60 V. As a rule of thumb, 
a safe value is usually approxi- 
mately twice the DC supply 
voltage. 

Test voltages 

The DC test voltages shown are 
measured with a 20 kI2/V instru- 
ment, unless otherwise specified. 

U, not V 

The international letter symbol 
'U' for voltage is often used 
instead of the ambiguous 'V'. 

'V' is normally reserved for Volts'. 
For instance: Ujj = 10 V, 
not Vf, = 10 V. 

Mains voltages 

No mams (power line) voltages 
are listed in Elektor circuits. It is 
assumed that our readers know 
what voltage is standard in their 
part of the world! 

Readers in countries that use 
60 Hz should note that Elektor 
circuits are designed for 50 Hz 
operation. This will not normally 
be a problem; however, in cases 
where the mains frequency is used 
for synchronisation some modifi- 
cation may be required. 

Technical services to readers 

• EPS service Many Elektor 
articles include a lay-out for a 
printed circuit board. Some - but 
not all — of these boards are avail- 
able ready-etched and predrilled. 
The 'EPS print service list' in the 
current issue always gives a com- 
plete list of available boards. 

• Technical queries Members o ( 
the technical staff are available to 
answer technical queries (relating 
to articles published in Elektor) 
by telephone on Mondays from 

14.00 to 16.30 Letters with 
technical queries should be 
addressed to Dept. TQ. Please 
enclose a stamped, self addressed 
envelope; readers outside U.K. 
please enclose an IRC instead of 
stamps 

• Missing link Any important 
modifications to, additions to, 
improvements on or corrections 
in Elektor circuits are generally 
listed under the heading 'Missing 
Link' at the earliest opportunity. 






contents 


elektor april 1978 — E 5 



It is generally agreed that, 
as far as sound quality is 
concerned, moving-coil 
pickup cartridges have 
the edge over their 
moving-magnet counter- 
parts. The moving coil 
preamp design presented 
here can be built for 
around one-tenth the 
cost of a comparable 
commercial unit. 

p. 4-02 


a ; — 

— i r~ 

i 

b 1 A A A A A A AA 

nr 

c nn 

i\i\i \j \j \i\i 



d km 

e pMI 

Ij/flf 1 1 1 1 INI 

If L r 


- Illllllll II II INI 



Low cost, high trans- 
mission rate and com- 
plete reliability were the 
design requirements for 
the cassette interface 
described in the 
following article. The 
interface makes very few 
demands on the sound 
quality of the recorder, 
and can transfer data at a 
rate of up to 1200 Baud. 

p. 4-20 



An orchestra suddenly 
begins to recite a passage 
of Shakespeare, an elec- 
tric guitar reads the news, 
the voice of a talker 
unexpectedly changes 
sex, a single voice sounds 
like a chorus — these are 
just a few of the amazing 
effects which can be 
obtained with a new elec- 
tronic instrument — the 
vocoder. This article 
explains the ins and outs 
of this fascinating new 
development in the field 
of electronic 'music'. 

p. 4-27 



selektor 4-01 

moving coil preamp 4-02 

electronic input selector 4-06 

elektornado 4-07 


The Elektornado is a high-fidelity amplifier offering ex- 
tremely good performance at moderate cost. The use of an 
1C for the input and driver stages reduces component count 
and allows an extremely compact construction. Two ampli 
fier channels can be accommodated on a single (small) 
printed circuit board either as a 2 x 50 W stereo amplifier or 
a 1 00 W mono bridge amplifier. 


stepped volume control 4-12 

real load resistors 4-13 

compander 4-14 


Until recently, companders were fairly complicated circuits. 
Now, however, they are available in the form of integrated 
circuits, one of which, the Exar XR 2216, is discussed in this 
article. This 1C can be used in a variety of applications such 
as amateur radio, PA systems, transcription of recorded 


material from disc to tape etc. 

stereo pan pot 4-17 

723 as a constant current source 4-18 


The pA 723 precision voltage regulator 1C is well known for 
its versatility, good line and load regulation, and low tem- 
perature coefficient. In addition to its many uses as a voltage 
regulator, it can also be used as a precision current regulator 


(constant current source). 

cassette interface 4-20 

car rip-off protection — W. Braun 4-26 

vocoders (1) — C. Chapman 4-27 



The uncommon design 
approach used in the 
moving coil preamp is 
reflected in the repetitive 
nature of the printed 
circuit board! 


formant — the elektor music synthesiser (10) . 

C. Chapman 

This final part of the Formant series completes the descrip- 
tion of the synthesiser by describing the COM (control and 
output module) and by giving an overall wiring diagram. 
Possibilities for further expansion of the system are also 
discussed. 

loudspeaker connections 

market 

tup-tun-dug-dus, ttl ics, cmos ics 

advertisers index 


4-33 


4-39 

4-40 

E-17 

E-21 


IV. 





selektor 


elektor april 1978 — 4-01 



General purpose active filter 

Due to the previous successes of the 
AF 1 00 Bi-Quad active filter and the 
AF1 20 Gyrator, National Semi- 
conductor has introduced two new 
standard ‘building block’ filters, the 
AF1 50 and AF1 5 1 . 

The AF1 50 is a high frequency version 
of the AF 1 00. Whereas the AF 1 00 is 
limited to an upper frequency of 
1 0 kHz and a center frequency - Q 
product of 50,000, the new AF150 has 
extended the upper frequency range to 
1 00 kHz and center frequency - Q 
products to 200,000. This has been 
accomplished by using high frequency 
operational amplifiers and laser trimmed 
resistors. 

Like its low frequency brother, the 
AF 1 50 is a Bi-Quad configuration and 
offers extremely low sensitivity to 
external component changes. It simul- 
taneously provides low pass, high pass 
and bandpass outputs while requiring 
only four external resistors to indepen- 
dently set the frequency, Q, and gain of 
the filter. In addition, by using an 
external operational amplifier, the 
LF356, all pass and notch filters can be 
formed. 

Another new device, the AF 151 is a 
dual Bi-Quad filter; that is, it provides 
two separate Bi-Quad active filters in 
one package. The performance of each 
filter section is identical to that of the 
AF100. This allows the user to easily 
design fourth-order filters in a single 
package. 

In fact, because the package also con- 
tains two uncommitted operational 
amplifiers that can be used as buffers, 
summing amplifiers or elsewhere in the 
system, a really clever designer could 
make use of these amplifiers to design 
up to eight order filters in a single 
package . For example, one AF 1 5 1 
could be used for the transmit filter and 
one AF 1 5 1 could be used as the receive 
filter in an asynchronous modem. It 
could easily be made to switch between 
originate and answer modes. 

It appears that most filter being syn- 
thesized using Bi-Quad filters are of 
fourth-order and higher. For this reason 
it was decided to put it all in one 
package to offer the user lower manu- 
facturing costs which are attendant with 


package insertion and smaller P.C. board 
space. 

‘The active filter revolution maybe 
analogous to the opamp revolution. As 
with an opamp, the user can simply put 
a few resistors around it and it does 
a specific job. It’s a general purpose 
device, easily tailored, and saves 
P.C. board space, design time and 
money’. 

There are thousands of uses for these 
filters. Once the designer establishes 
the center frequency, Qand gain 
requirements, a complicated sixth order 
filter design will normally take less than 
half an hour. 

National Semiconductor GmbH 
Industrie stra fie 1 0 
D-8080 Fiirstenfeldbruck 
West Germany 

(295 S) 


Electronic shredder 

The electronically sensing Fordishred 
1800 represents a major advance in 
shredder technology. In addition to a 
power overload cut-out sensor, the 1 800 
features a unique electronic sensing 
device for paper load control. If the 
shredder is overloaded, the electronic 
sensor detects this before jamming 
occurs. The offending material is then 
rejected for separation and rein- 
troduction. This means the operator can 
leave the machine unattended, secure in 
the knowledge that it will deal safely 
and efficiently with any prospective 
overload . 

The Fordishred 1800 is precision- 
engineered and robustly built to take 
the strain of constant heavy duty 
shredding in commercial or industrial 
environments. Suitable for centralised 
operation, the powerful 1800 has an 
1 8%” wide throat and a voracious 
appetite, shredding cardboard and paper 
waste into V*” strips at the rate of % ton 
per hour of continuously fed paper. 
Operating from a standard 1 3 amp 
power point, and working at a speed of 



60 ft per minute, the machine will shred 
a complete file of some 50 sheets 
including staples, pins and paper 
clips — in one pass! Destroyed material 
is simply ejected beneath the machine 
into a polythene sack. The sack is 
securely mounted on runners and, when 
full, can be slid out and replaced with a 
new one in seconds. 

The 1 800 is also an excellent machine 
for turning waste into profit by recycling 
waste paper to produce good quality 
packaging material for internal use or 
resale. 

The machine is mounted on a rigidly 
constructed stand with four rubber- 
wheeled castors for complete mobility : 
the front two castors are fitted with 
brakes so that the shredder will remain 
completely stable whilst in use. To aid 
the operator further, the 1 800 can be 
supplied with either a feeding shelf and 
side table for storing material prior to 
shredding or with a large work shelf for 
sorting loads. 

The control panel, which faces the 
operator while the machine is in use, 
has well-spaced out, colour coded, 
‘forward’ ‘reverse’ and ‘stop’ buttons. 
For added safety a master key is 
required to switch on the power. 
Conforming to BS 4644 and BS 3861 
specifications for electrical and 
mechanical standards, the 1 800 is 
powered by a 1V4 hp motor and operates 
on standard single phase 13 amp 
electricity supply. It measures 
790 mm x 590 mm x 1092 mm high 
including stand (31” x 23” x 43”). The 
price is £ 1400 excluding VAT. 

Fordigraph Division of Ofrex Limited, 
Ofrex House, Stephen Street, 

London WlA 1EA, England. 

(289 S) 


mini-selektor 

Semiconductors take over from 
TWTs 

The Austrian Broadcasting Corporation 
(ORF) has replaced a total of ten 
travelling-wave-tube (TWT) amplifiers in 
five major transmitting stations by solid- 
state amplifiers of the type VD 1 10 
from Rohde & Schwarz. Both technical 
and economic considerations played a 
role in the decision to reequip the low- 
power stages of the Band IV/V TV 
transmitters for the second program. 

At about 400 VA each, the power 
consumption of the solid-state amplifiers 
is low compared with the 3 kVA of a 
TWT stage. The costs of the transition 
will be covered in about two years by 
the elimination of the tube replacement 
costs. With a gain of > 30 dB and a 
bandwidth of 470 to 860 MHz, the 
VD 1 10 provides the same functional 
performance as its predecessor. 

(291 S) 


4-02 — elektor april 1978 


moving coil preamp 



In these days of laser scanners and micro- 
computers the mechanical process of a 
stylus being wiggled about in grooves, 
which have been scratched out of plastic 
to resemble the shape of sound waves, 
seems an incredibly crude and old- 
fashioned system. If it appears remark- 
able that such a process has continued 
in widespread use right up to the present 
day, it is even more remarkable that 
what could be described as such a con- 
ceptually primitive method allows sound 
to be reproduced with such startlingly 
good quality. 

Innummerable constructional improve- 
ments have ensured that the record 
player has kept pace with the ever- 
increasing demands placed upon the fid- 
elity of audio equipment. Better motors, 
improved drive systems, lighter and 
higher compliance pickup arms, the in- 
troduction of anti-skating compensation, 
and last but not least, a considerable im- 
provement in the performance of pickup 
cartridges are all steps in the evolution 
of the old acoustic gramophone into the 
modern hi-fi record player. 

This process of continual improvement 
is still evident, although it is now con- 
siderably less dramatic than in the past. 
The comparatively recent advent of 
direct-drive and crystal-controlled turn- 
tables are ample illustration. However as 
far as the domestic user is concerned, 
the latter innovation for example, can 
be counted among the category of snob- 
value ‘improvements’ which, although 
measurable, are audibly undetectable. 
And it is a somewhat regrettable fact 
that a number of similar developments 
are designed more to stimulate sales 
than improve the user’s appreciation of 
the reproduced sound. 

Nonetheless, leaving aside such commer- 
cially-inspired innovations, there are still 
many areas where useful improvements 
in the audio chain can be made, and one 
which has - justifiably - received recent 
attention is the pickup cartridge. 

A commendable development in this 
respect has been the trend to view the 
cartridge and pickup arm as a single unit, 
recognising the fact that one cannot as- 
sess the performance of the one without 
taking the other into account as well. 
A happy consequence of this has been 
the realisation that an extremely high 


It is generally agreed that, as far as 
sound quality is concerned, 
moving-coil pickup cartridges have 
the edge over their moving-magnet 
counterparts. The prices of moving- 
coil and moving-magnet cartridges 
of comparable performance are 
similar, but unfortunately the low 
output voltage of moving-coil 
cartridges necessitates the use of a 
step-up transformer or 
preamplifier which can cost more 
than the cartridge itself. 

For this reason alone the 
preamplifier design presented here 
should be welcomed by those who 
like to construct their own hi-fi 
equipment, as it can be built for 
around one-tenth the cost of a 
comparable commercial unit. 



compliance and an almost impracticably 
low tracking force are not necessarily a 
prerequisite for top class cartridges. 
Possibly these considerations have gained 
ground due to the increasing acceptance 
of the view among designers and re- 
viewers of hi-fi equipment that, in ad- 
dition to all the sophisticated electronic 
test equipment currently available, we 
possess two highly advanced but ex- 
tremely inexpensive measuring devices 
in the form of our ears! This is a trend 
which cannot be welcomed too strongly, 
since ultimately, the assessment of any 
link in the audio chain must be deter- 
mined not by specifications, distortion 
figures, and the like, but by the subjec- 
tive, if informed, response of the 
listener. 

Moving coil cartridges 

The above tendency provides a partial 
explanation for the recent increase in 
the popularity of moving coil cartridges. 
Although in the past they have been 
accused of poor tracking ability, and 
suffered from the disadvantage that 
they have to be returned to the manu- 
facturer for stylus replacement, as well 
as being comparatively expensive, there 
has never been any doubt about the mu- 
sical quality of moving coil pickups. 


Specification 

Frequency response: 

7 Hz to 80 kHz, 

Voltage gain: 

+0, —3 dB. 

33,5 dB (l) 

Input impedance: 

75 fl (') 

Output impedance: 

< 100 a 

Recommended load 

47 k 

impedance: 

Maximum input voltage: 

23 mV 

Total harmonic distortion 
for Uj n = 4 mV: 

< 0,05% 

Channel separation: 

< 60 dB 

Signal -to -noise ratio: 

< 68 dB I 2 ) 

Supply voltage: 

10. . . 20 V 

Current consumption 
(stereo version): 

100mA 

Notes 

I 1 ) Adjustable 

( 2 ) Reference level is the output voltage 

produced using an Ortofon MC-20 cartridge 

tracking at 10 cm/sec. 




moving coil preamp 


elektor april 1978 — 4-03 



Figure 1. Block diagram showing the principle 
of a noise-cancelling amplifier. 

Figure 2. Complete circuit of one channel of 
the moving coil pickup preamplifier. 


While top-class cartridges of both 
moving-magnet and moving-coil types 
are capable of excellent performance, 
the sound of a moving-coil cartridge 
possesses a clarity and transparency not 
obtained from moving magnet types. 
Thus many reviewers were inclined to 
have mixed feelings about this type of 
cartridge, since listening tests often gave 
much better results than could be 
expected on the basis of measured per- 
formance. 

The question which no doubt most pro- 
spective buyers will ask, namely whether 
moving coil pickups are better than 
moving magnet types, fortunately does 
not fall within our brief. Indeed this is a 
question to which even reviewers of hi-fi 
equipment cannot really be expected to 
provide a generally valid reply, since the 
sound quality will be assessed differently 
by each reviewer. 

A distinctive feature of moving coil 
pickups is the exceptionally direct and 
clear reproduction, which tends to dis- 
tinguish them from their moving magnet 
colleagues. However there are top- 
quality moving magnet cartridges which 
often seem to possess just the right 
character for particular listeners, so that 
the serious audiophile should always 
take the trouble to compare different 
types of cartridges before making a pur- 
chase. 

Unfortunately however, an effective 
comparison of different types of 
pickup is not always possible, since a 
moving coil cartridge produces an out- 
put voltage that is only a fraction of 
that produced by a moving magnet type. 
Thus to the not inconsiderable price of 
a moving coil cartridge has to be added 
the cost of a step-up transformer or — 
preferably, in view of its higher fidelity 
and lower sensitivity to hum — that of a 
special preamplifier. A suitable trans- 
former will cost around £ 15, whilst a 
preamplifier could cost anything be- 
tween £ 50 and £ 80 — sufficient reason 
for many prospective users to settle for 
a moving magnet cartridge. 


It is clear therefore that a powerful ar- 
gument exists for building a suitable 
preamp oneself, thereby obviating the 
need to sacrifice one’s musical discrimi- 
nation for reasons of cost. 

Preamplifier 

Building a good preamplifier for very- 
low signal levels is no easy matter, and 
the output signal of a moving coil 
pickup is extremely small indeed. 
Ortofon cartridges (which we have 
taken as a reference, since they have 
about the lowest output signal and since 
Ortofon is the only manufacturer of 
pickups who also produces a separate 
preamplifier) deliver approximately 
70 /TV per channel at an output im- 
pedance of 2 SI. Thus a gain of around 
50 would be required to boost this to 
the output level of an average moving 
magnet cartridge. It is clearly a tricky 
task amplifying such minute signals 
whilst maintaining an acceptable signal- 
to-noise ratio. By acceptable we mean a 
figure of at least 65 dB. 

There are only a limited number of 
possibilities for the design of a suitably 
linear and low-noise preamp. One could 
look for an ultra-low-noise semiconduc- 
tor. However, such a transistor would 
almost certainly be exorbitantly expens- 
ive, whilst availability would also present 
a thorny problem. Thus this approach 
does not seem very promising for a do- 
it-yourself type of project. 

The alternative is to construct a simple 
but inherently low-noise amplifier stage 
and then see which readily available 
transistors give the best noise figure. 
Once this has been ascertained, the cir- 
cuit is optimised for this particular tran- 
sistor. Then a number of these amplifier 
stages are connected in parallel, as 
shown in figure 1 . This trick was already 
explained in the article on the noise can- 
celling preamp which was published in 
last year’s Summer Circuits issue (Elektor 
27/28, circuit 75). If n identical ampli- 
fiers are connected in parallel then as far 
as voltage gain is concerned they will 



6V 

50mA 


0 











4-04 — elektor april 1978 


moving coil preamp 


3 

D1 


1N4001 



function as a single amplifier, since each 
is fed with the same input signal, and 
each has the same gain. The outputs of 
all the amplifiers are thus equal and in 
phase. The noise voltages generated by 
the individual amplifiers, however, are 
random and, in mathematical terms, are 
uncorrelated with one another. Partial 
cancellation of the noise voltages will 
therefore occur at the amplifiers’ com- 
mon output. The result is that the signal- 
to noise ratio of the output signal is ef- 
fectively increased by a factor of V n > 
where n is the number of amplifier stages 
connected in parallel. 

In the case of this circuit, which contains 
eight amplifier stages, this means an im- 
provement in the signal to noise ratio of 
9 dB. To connect more than 8 stages in 
parallel is not considered worthwhile, 
since with such an arrangement the law 
of diminishing returns is applicable: to 
obtain a further (audible) improvement 
of 3 dB would require eight extra stages, 
and so on. 

The circuit 

The most obvious feature in the circuit 
diagram shown in figure 2 is the chain 

of amplifiers T1 T8. Although this 

arrangement may offend the aesthetic 
sensibilities of some readers, the result- 
ant signal-to-noise ratio (> 68 dB ! ^ tes- 
tifies to its efficacy. 

After some experiment, a reasonably 
cheap and commonly available transis- 
tor was found for the input stages: the 
BP 494. It may seem a surprising choice 
at first sight, since this transistor is nor- 
mally used in high frequency circuits. 
However the BF 494 is much more ac- 
customed to handling very small input 
signals, and in fact proved more suited 
to this application than the members of, 
e.g., the BC family of transistors. 
Additional voltage gain is provided by 
T9, and emitter-follower T10 acts as a 
low impedance output buffer capable of 
driving the relatively low-impedance 
feedback loop as well as the re- 



Figure 3. Stabilised power supply for the 
preamplifier which requires a 10 to 20 V 
input and provides a regulated 6 V output. 

Figure 4. Printed circuit and component lay- 
out for a stereo preamp plus power supply 
(EPS 9911). 


Parts list to figure 2 and 4. 

Note: for stereo version, two of 
each required. 

Resistors: 

(preferably metal film) 

R1 = 82 ft 
R2 . . . R9 = 18 k 
RIO. . . R17 = 6k8 
R18 = 68 k 
R19 = 270 k 
R20.R23 = 470 n 
R21 = 47 ft 
R22 = 1 n 

Capacitors: 

Cl =22n 

C2 . . . C9 = 470 p/3 V 
CIO . . . C17 = 4p7/10 V 
C18 = 4n7 
C19 = 33 n 
C20 = 47 p/10 V 
C21 = 470 p/10 V 
C22 = 100 n 

Semiconductors: 

T1 . . . T8 = BF 494 (BF 194, 

BF 195. BF 495) 

T9 = BC560C, BC559C, 

BC 1 79C or equ. 

T10 = BC547B, BC 107B or equ. 


Parts list to figure 3 and 4. 


Note: only one of each required. 

Resistors: 

R24 = lk5 
R25 = 8k 2 
R26 = lk2 
R27 = 1 k 
R28 = 100 n 
R29 = 8n2 

Capacitors: 

C23 = 100 p/10 V 
C24 = 1 n 
C25 = 4p7/40 V 

Semiconductors: 

IC1 = 723 (pA723, LM723, etc.) 
Til = BD 241 (fitted with heat 
sink) 

D1.D2 = 1N4001 


commended output load. R21, R22 and 
Cl 9 form the negative feedback loop, 
the mid-band gain being given by the 

R 21 

equation A = 1 + - — . 

R22 

With the component values given, the 
mid-band gain is exactly 48. This is 
almost exactly the figure needed to 
boost the output voltage of an Ortofon 
moving coil pickup to the level of 
approx. 3.5 mV, which is the average 
output level of a moving magnet car- 
tridge. 

If one has a moving coil cartridge with a 
greater output voltage (Denon cartridges 
have an output roughly 4 times greater 
than average moving coil pickups), then 
the value of R22 can be increased to 




moving coil preamp 


elektor april 1978 — 4-05 



2.2 £2. This reduces the gain by more 
than half, so that there is no danger of 
overloading the disc input of the suc- 
ceeding audio amplifier. 

The input impedance of the amplifier is 
fairly low, and is largely determined by 
the value of R 1 . With the value given in 
the circuit diagram ( R 1 = 82 f2) the in- 
put impedance exactly coincides with 
the recommended load impedance for 
the Ortofon moving coil pickup of 75 f2. 
Other input impedances can be obtained 
by altering the value of R1 accordingly. 

Power supply 

The stereo version of the preamp 
requires a supply voltage of 6 V at 
100 mA. This is supplied by a 723 IC 
voltage regulator with external transis- 


tor, as shown in figure 3. The regulator 
circuit requires an input of between 10 
and 20 V at 100 mA. It may be possible 
to obtain this voltage from some existing 
piece of equipment such as an audio 
amplifier or preamp, but if such a voltage 
is not available then it must be provided 
by a separate transformer, bridge recti- 
fier and smoothing capacitor. The trans- 
former should have an RMS output volt- 
age between 9 and 15 V at 160 mA and 
the bridge rectifier and capacitor should 
be rated at 30 V 100 mA and 220 n 
(minimum) 25 V respectively. 

Construction 

Figure 4 shows a printed circuit board 
which will acommodate a stereo version 
of the preamp plus the supply stabiliser. 


It goes without saying that the compo- 
nents used in the construction must all 
be of the highest quality, otherwise the 
s/n ratio may be degraded. Metal-film 
and metal-oxide types are to be pre- 
ferred for the resistors, and tantalum 
types are preferred for the capacitors. 
The transistors should have the mark of 
a reputable manufacturer, and if possible 
the 1 6 input devices ahould all be from 
the same production batch. 

As the signal levels in the circuit are ex- 
tremely low, a great deal of care must 
be taken in the construction. The printed 
circuit board must be housed in a 
totally screened (metal) case, and all 
signal leads should be of low-noise 
screened cable. To avoid earth loops the 
input socket(s) should be insulated 




4-06 — elektor april 1978 moving coil preamp 


electronic input selector 




electronic 
input selector, 



Figure 5. Photograph of the completed proto- 
type, with the cover removed. 

Figure 6. Another view of the prototype with 
the cover in place. In view of the low input 
signal levels, gold-plated phono connectors are 
recommended for the input sockets. 


from the ease. 

If the transformer must be mounted in 
the same housing as the preamp then it 
should be in a screened enclosure of its 
own within the main housing, in order 
to minimise hum pickup, and should be 
as far away as possible from the input of 
the preamp. However, in general it is 
recommended that the transformer is 
kept well away from the preamp! 

Testing 

The preamp should work immediately 
when it is switched on and a suitable sig- 
nal is fed in. In the unlikely event of a 
fault occurring, however, the test point 
voltage shown in figure 2 should be 
checked. Furthermore the collector 
voltages of transistors T1 to T8 should 
be approximately IV. If the collector 
voltage of one (or more) transistors is 
significantly different from the rest then 
it is best to replace the offending device, 
since it will probably exhibit different 
characteristics from the other devices, 
which could have a detrimental effect 
on the s/n ratio. 

Apart from this, provided care is taken 
in the construction, no problems should 
be encountered and the preamp should 
deliver a performance which compares 
favourably with that of commercial de- 
signs costing many times more. K 


In most audio amplifiers the input 
selector switch is mounted on the front 
panel of the equipment, whilst the input 
sockets are mounted on the back panel. 
This means that the input signal leads 
must be routed all the way from the 
back panel to the front panel before 
going off to the actual amplifier input, 
thus increasing the possibility of hum 
and noise pickup crosstalk. 

The transistor input selector described 
here switches the signals at the rear of 
the amplifier close to the input sockets. 
Switching is still controlled from the 
front panel switch, but audio signals no 
longer flow through it. 

One channel of the selector is shown in 
figure 1. Each input is fed to an emitter 
follower whose base bias voltage is 
obtained from the selector switch SI. 
When a particular input is selected by 
SI then the appropriate transistor 
receives a base bias voltage and is able 
to pass the input signal. The bases of all 
the other transistors are pulled down to 
ground by resistors Rg and are thus cut 
off. 

Since SI supplies only a DC bias voltage 
to the selector circuit the length of lead 
between SI and the input selector is 


unimportant. An additional bonus is 
that the transistors function as im- 
pedance converters. The low output 
impedance means that there is no 
restriction on the length of lead 
between the output of the selector 
circuit and the input of the amplifier. 
Switching clicks are also suppressed 
since capacitors C2 cause the base volt- 
ages to die away smoothly rather than 
switching the transistors abruptly as SI 
is operated. 

The circuit can be extended to any 
number of inputs and any number of 
channels simply by adding an extra 
transistor for each additional input and 
duplicating the total selector circuit for 
each additional channel required. M 


Specification 


Impedance of each input 

100 k 

Output impedance 

< 1 k 

Maximum input voltage 

= 1 V RMS 


(3 V peak 


to peak) 

Gain 

= 1 (OdB) 








elektornado 


elaktor april 1978 - 4-07 



There has been much debate in hi-fi 
circles about the necessity for high 
amplifier output powers, with some 
maintaining that a high output power is 
an absolute necessity for undistorted 
handling of programme peaks, and 
others maintaining that high-power 
amplifiers are just a status symbol. Be 
that as it may, there is no doubt that 
for many applications such as disco 
work, or situations where extremely 
inefficient loudspeakers are being used, 
a high power output is a definite 
advantage. With its choice of SOW or 
100 W maximum output power, the 
Elektornado should certainly satisfy 
most requirements. 

A high output power either entails the 
use of a high supply voltage or the use 
of a bridge output stage. A bridge 
configuration was chosen for the 
Elektornado for several reasons: 

1. It allows relatively inexpensive out- 
put devices to be used and avoids the 
necessity for expensive high-voltage 
(> 60 V) devices. 

2. Each half of a bridge amplifier can 
be used as an independent, lower power 
amplifier. 

A bridge configuration entails the con- 
struction of two virtually complete 
amplifiers for each channel, so some 
way had to be found of reducing com- 
ponent count. Fortunately, the input 
and driver stages of the amplifier can be 
replaced by a single IC which has 
recently been introduced, the LM391. 
In the past, integrated circuits have not 
been very suitable for hi-fi applications 
due to limitations of bandwidth, distor- 
tion, noise and operating voltage. The 
LM391, however, suffers from none of 
these disadvantages. 

The circuit 

The complete circuit of one channel of 
the amplifier, including the equivalent 
internal circuit of the LM 391 , is shown 
in figure 1 . The IC replaces all the input 
and pre-driver stages of the amplifier, 
the only parts of the circuit using dis- 
crete transistors being the driver and 
output stages. 

The input stage of the IC consists of a 
differential amplifier (Tq, Th) and a 
current mirror (Tq, Tp), which forms 
the collector loads for the differential 


The Elektornado is a high-fidelity 
amplifier offering extremely good 
performance at moderate cost. 

The use of an IC for the input and 
driver stages reduces component 
count and allows an extremely 
compact construction. Two 
amplifier channels can be 
accommodated on a single (small) 
printed circuit board either as a 
2 x 50 W stereo amplifier or a 
100 W mono bridge amplifier. 


Table 1 measured with + 30 V supply. 

Maximum output power: 
stereo 2 x 45 W into 4 ohms 

2 x 50 W into 8 ohms 
mono (bridge) 100 W into 8 ohms (4 ohm 
load not recommended as 
current limiting occurs at 
45 W) 

Frequency response: 

6 Hz to 30 kHz (-3 dB) 

Total harmonic distortion: 

0.1% 40 Hz <f < 10 kHz 
(see also figure 6) 


stage. The signal from the collector of 
Th is fed to a cascode stage (To, T{q), 
which has a very high gain, and thence 
to the output stages of the IC. 

The driver and output stages of the 
amplifier consist of two discrete tran- 
sistor pairs T1/T3 and T2/T4, the 
quiescent current of the output stage 
being set by the collector/emitter volt- 
age of transistor Tk, which is varied by 
adjusting the base bias by means of P 1 . 
To avoid distortion caused by slew-rate 
limiting (slope overload), care has been 
taken in the design of the feedback and 
compensation networks, and additional 
protection is provided in the form of an 
input filter R15/C11, which limits the 
slew-rate of the input signal. However, 
this does not have a detrimental effect 
on the normal frequency response, 
which begins to roll off at about 30 kHz. 
The closed-loop gain of the amplifier 
is determined by the feedback network 
R5, R1 and Cl. At frequencies where 
the reactance of Cl is small the gain is 
given by: 


Ay = 


Uout 

uI7 



^ 22 . 


At low frequencies the increased reac- 
tance of Cl in series with R1 causes the 
gain to roll off to unity for DC signals. 
Amongst other things this avoids any 
DC offset problems which might result 
from a high DC gain. With the com- 
ponent values shown the voltage gain 
is approximately 20 (26 dB), which 
means that the input sensitivity for 
full output voltage swing is about 1 volt. 
This should make the circuit suitable for 
use with most modern preamps. 


Circuit protection 

Several protection circuits are incorpor- 
ated into the design to prevent damage 
to the output transistors under various 
fault conditions. 

Inductor LI, which is wound on R18, 
protects the output stage when oper- 
ating into capacitive loads. 

Diodes D1 and D2 provide brute-force 
protection against any transients that 
might be produced by an inductive load 
by clamping the maximum output volt- 
age excursion to ± Ut>. 

A number of sophisticated protection 
circuits exist within the IC itself. Should 




4-08 — elektor april 1978 


elektornado 


____ 

LM 391 

15... 30V 




the output current of the amplifier rise 
above about 4 amps peak the voltage 
dropped across R12 or R13 will cause 
transistors Tl or Tm to turn on, thus 
limiting the output current. 

Thermal protection of the output tran- 
sistors may also be provided as an op- 
tional extra if desired. A negative tem- 
perature coefficient thermistor, which 
is in thermal contact with the output 
transistor heatsink, may be connected 
between pin 14 of the IC and ground. 
Current will flow through this thermis- 
tor via the two base resistors of Ta- As 
the temperature increases and the resist- 
ance of the thermistor falls this current 
will increase until the voltage drop 
across the 5 k resistor is sufficient to 
turn on Ta- This will shut down current 
sources Tg, Tc anc * Tp and cut off the 
drive to the output stage. 

A resistor may need to be included in 
series with the thermistor to limit the 


maximum current out of pin 14 to 
1 mA, and the thermistor value should 
be chosen such that the current out of 
pin 1 4 will be about 1 00 /uA at the 
desired cutoff temperature. 

Two-channel amplifier 

Figure 3 shows the complete circuit of a 
two-channel amplifier. In this case, for 
simplicity, the internal circuit of the 
LM 39 1 is not shown. With a ± 30V sup- 
ply each channel of the amplifier will 
deliver 50 W into an 8 ohm load, or 
45 W into a 4 ohm load. By connecting 
a resistor, R x , between the output of 
one channel and the inverting input of 
the other channel (non-inverting input 
grounded) the two channels can be 
made to function as a mono bridge 
amplifier, with the loudspeaker connec- 
ted as shown dotted. Note that in this 
configuration both ends of the loud- 


speaker are floating! 

Theoretically, the maximum output 
power that can be obtained in the 
bridge mode is four times that obtained 
in the normal configuration. However, 
this would place great stress upon the 
output transistors and would require 
much more massive heatsinks and an 
extremely ‘beefy’ power supply. The 
maximum output power into an 8 ohm 
load in the bridge configuration is there- 
fore restricted to 1 00 W by current 
limiting. Operation into a 4 ohm load in 
the bridge configuration is not rec- 
ommended as current limiting will re- 
strict the maximum output power to 
about 45 W. 

Printed circuit board 

The printed circuit board and compo- 
nent layout for the Elektornado are 
given in figure 4, and it will be seen that 









elektornado 


elektor april 1978 — 4-09 



two identical channels are mounted on a 
single board to facilitate operation in 
the bridge mode. If this configuration is 
required then R x is soldered into place 
and the input of the left channel is 
grounded. If the 2 x 50 W stereo version 
is required then R x is omitted. 

LI consists of 20 turns of 0.9 mm 
(20SWG) enamelled copper wire, wound 
on the body of resistor R18. 

The driver and output transistors are, 
of course, mounted external to the 
board on heatsinks, which should have a 
thermal resistance of less than 1.5°C per 
watt and should be mounted with the 
fins running vertically to give a chimney 
effect which will aid cooling. Painting 
the heatsinks matt black also improves 
cooling. 

Wiring 

To avoid problems of instability, earth 


loops etc. the wiring layout shown in 
figure 5 should be followed. For clarity 
the driver and output transistors are not 
shown in this diagram. A simple, un- 
stabilised power supply between ± 1 5 V 
and ± 30 V is quite adequate for the 
amplifier, although the maximum out- 
put power will only be obtained with 
the higher supply voltage. Care should 
be taken to ensure that the off-load 
voltage of the power supply is no 
greater than ± 30 V, otherwise there is a 
danger of damaging the IC or output 
transistors. The 2 x 20 V RMS sec- 
ondary rating of the transformer should 
be considered as an absolute maximum, 
as this will allow for a +10% variation in 
mains voltage. 

Setting quiescent current 

Before applying power to the amplifier 
PI and PI’ should be turned fully to the 


Figure 1. Circuit of one channel of the Elek 
tornado, showing the internal circuit of the 
LM 391 IC. 

Figure 2. Pinning of the driver and output 
transistors (all bottom view). 

Figure 3. Complete circuit of the 50 W per 
channel/100 W mono amplifier. 

Table 1. Principal specifications of the 
Elektornado amplifier. 



4-10 — elektor april 1978 


elektornado 


Parts list to figure 5. 

Resistors: 

R1,R1',R4,R4’ = 4k7 
R2,R2’,R5,R5', 
R6,R6',R9,R9',R X = 100 k 
R3,R3' = 47 k 
R7,R7',R8,R8',R15,R15', 
R16,R16’,R17,R17' = 1 k 
RIO, RIO', R1 1 ,R1 1 ’ = 100 n 
R12,R12',R1 3,R13' = 0.15 17/3 W 
R14.R14' = 10 Sll 1 W (carbon 
film resistor) 

R18.R18' = 1 n/1 W (carbon 
film resistor) 

PI, PI’ = 10 k preset 

Capacitors: 

C1,C1' = 4p7/16V 
C2,C2‘ = 1 p/63 V 
C3,C3',C6,C6' = 1 0 p/63 V 
C4,C4',C7,C7' = 4p7 
C5,C5',C10,C10',C13,C13', 
C14,C14'= 100 n 
C8,C8',C9,C9',C1 1 ,C1 1 ' = 1 n 
C12,C12' = 47 n 

Semiconductors: 

IC1,IC1'= LM 391 -60 or 
LM 391-80 
T1 ,T 1 ' = BD 139 
T2,T2'= BD 140 
T3,T3' = TIP 2955 or MJE 2955 
T4,T4' = TIP 3055 or MJE 3055 
D1,D1',D2,D2’= 1N4002 


Figure 4. Printed circuit board and com- 
ponent layout for the Elektornado (EPS 9874). 

Figure 5. Wiring diagram for the Elektornado 
(driver and output transistors not shown). 
Figure 5a: stereo version; figure 5b: bridge 
version. 


Figure 6. Total harmonic distortion versus 
frequency graph for the Elektornado. 










elektor april 1978 — 4-1 ^ 


elektornado 


B40 C5000 


sec 

2 * 20V 
4A 


C L .4700p 40V 


N.B. For clarity, the output transistors are not 
shown. 


B40 C5000 


Cj^ 4700*/ 40V 


N.B. For clarity, the output transistors are not 
shown. 


tion is below 0.1% over the entire audio 
spectrum, and over the important mid- 
band frequencies is less than 0.02%. 
Other important parameters of the 
amplifier are listed in table 1. 

As mentioned previously, the input sen- 
sitivity for full output is 1 V RMS. 
which should be suitable for most pre- 
amplifiers. However, if this sensitivity is 
insufficient the gain of the amplifier 
may be increased simply by changing 
the values of R1 and R5 (decreasing R1 
and/or increasing R5). 

The high output power and excellent 
specifications of the Elektornado, 
together with its versatility, should en- 
sure that it will prove the right answer 
for a great number of amplifier appli- 
cations. M 


left. 

A multimeter set to the 100 mA range is 
then connected in the positive or nega- 
tive supply lead to the left channel, and 
PI is adjusted to give a current of be- 
tween 50 and 100 mA. The procedure is 
then repeated for the right channel. 

If the amplifier should exhibit any tend- 
ency to instability (this may manifest it- 
self as an excessively large and uncon- 
trollable quiescent current) this can be 
cured by increasing the values of C4 and 
C7, keeping them of equal value. 


Conclusion 

The specifications of the Elektornado 
can safely be called excellent. As can be 
seen from figure 6 the harmonic distor- 












4-12 — elektor april 1978 


stepped volume control 


stepped 

volume control 


Conventional rotary or slider potentio- 
meters suffer from several disadvantages 
when used as volume controls in an 
audio system. The ganged, logarithmic 
potentiometers which are frequently 
employed in stereo amplifiers fre- 
quently suffer from poor matching of 
the two channels, so that the relative 
signal levels or balance of the left- and 
right channels vary as the control is 
operated. Carbon potentiometers also 
have a relatively limited life and soon 
become noisy in operation. 

One solution to these problems is to use 
a stepped volume control consisting of a 
switched, resistive potential divider, as 
shown in figure 1 . This circuit has sev- 
eral advantages over a conventional 
potentiometer. 

— matching between channels is deter- 
mined solely by resistor tolerances 
(5% tolerance should be adequate for 
most applications) 

— the control can be made to have any 
desired ‘law’ by suitable choice of 
resistor values 

— within reason, any number of chan- 
nels can be catered for by using a 
switch with more wafers 

a long life is obtained, provided a 
reasonable quality switch is used. 

The degree of attenuation produced for 
a particular setting of the control is 
given by attenuation = 20 log (R r : Rt)dB, 
where Rt is the total resistance of the 
potential divider chain and R r is the 
remaining resistance between a particu- 
lar switch position and ground. The 
value of individual resistors connected 
between two adjacent positions of the 
switch is obviously obtained by sub- 
tracting two adjacent values of R r . 

For a volume control a logarithmic law 
is desirable, which means that the dif- 
ference in attenuation between any two 
adjacent settings of the control must be 
a constant number of dB. Table 1 shows 
the values of R r required for 1 dB steps 
of attenuation from 0 to -60 dB for an 
Rt value of 100 k (plus an extra step for 
infinite attenuation). Obviously a prac- 
tical volume control cannot have this 
number of steps, as this would require a 
62-way switch. On the other hand, the 
number of switch positions must not be 
too small, as this will not give suf- 
ficiently fine control. 



Table 1 


dB 

R r ( R t = 100.000 n ) 

dB 

R r IR , = 100.000 n ) 

0 

100.000 

-31 

2.818 

-1 

89.125 

-32 

2.512 

-2 

79.794 

-33 

2.239 

-3 

70.794 

-34 

1.995 

-4 

63.095 

-35 

1.778 

-5 

56.234 

-36 

1.585 

-6 

50.118 

-37 

1.413 

-7 

44.668 

-38 

1.259 

-8 

39.810 

-39 

1.122 

-9 

35.481 

-40 

1.000 

-10 

31.622 

—41 

891 

-11 

28.184 

-42 

794 

-12 

25.119 

-43 

708 

-13 

22.387 

—44 

631 

-14 

19.952 

-45 

562 

-15 

17.783 

-46 

502 

-16 

15.849 

-47 

447 

-17 

14.125 

-48 

398 

-18 

12.589 

-49 

355 

-19 

11.220 

-50 

316 

-20 

10.000 

-51 

282 

-21 

8.913 

-52 

251 

-22 

7.943 

-53 

224 

-23 

7.079 

-54 

200 

-24 

6.310 

-55 

178 

-25 

5.623 

-56 

158 

-26 

5.012 

-57 

141 

-27 

4.467 

-58 

126 

-28 

3.981 

-59 

112 

-29 

3.548 

-60 

100 

-30 

3.162 

— oo 

0 



stepped volume control 


real load resistors elektor april 1978 - 4-13 


Table 2 






1 

2 

3 

4 

5 

6 

0 

100.000 

29.206 

29.200 

99.972 

0 

-3 

70.794 


(27k+2k2) 

70.772 

-3.0 



20.676 

20.600 



-6 

50.118 


(1 5k+5k6) 

50.172 

-6.0 



1 4.637 

14.700 



-9 

35.481 


(10k+4k7) 

35.472 

-9.0 



10.362 

10.390 



-12 

25.119 


(10k+390n) 

25.082 

-12.0 



7.336 

7.360 



-15 

17.783 


<6k8+560n) 

17.722 

-15.0 



5.194 

5.170 



-18 

12.589 


(4k7+470n) 

12.552 

-18.0 



3.676 

3.630 



-21 

8.913 


<3k3+330n) 

8.922 

-21.0 



2.603 

2.620 



-24 

6.310 


I1k8+820n) 

6.302 

-24.0 



1.843 

1.847 



-27 

4.467 


(1 k8+47D) 

4.455 

-27.0 



1.305 

1.300 



-30 

3.162 


I1k2+100n) 

3.155 

-30.0 



923 

920 



-33 

2.239 


(82on+ioon) 

2.235 

-33.0 



654 

642 



-36 

1.585 


(560n+82n) 

1.593 

-36.0 



463 

470 



-39 

1.122 


(470S7) 

1.123 

-39.0 



328 

330 



-42 

794 


(330J7) 

793 

-42.0 



232 

232 



-45 

562 


(22017+1 2ft) 

561 

-45.0 



164 

164 



-48 

398 


(82n+82J7) 

397 

-48.0 



116 

120 



-51 

282 


112017) 

277 

-51.1 



82 

82 



-54 

200 


(82J7) 

195 

-54.2 



59 

56 



-57 

141 


(56fi) 

139 

-57.1 



41 

39 



-60 

100 


(3917) 

100 

-60.0 



100 

100 



_oo 

0 


(1 0017) 

0 

_ oo 


A reasonable choice of attenuation step 
is 3 dB. This gives sufficiently fine con- 
trol, yet allows 60 dB of attenuation to 
be achieved in 2 1 steps. Allowing an 
extra step for the zero (infinite attenu- 
ation) position means that 22 ways are 
required in all. 

The resistance values for a 22 position 
control are given in table 2. Column 1 
lists the required attenuation in dB for 
each switch position. Column 2 lists the 
corresponding values of R r . Column 3 
lists the resistor values required between 
the switch positions. Column 4 lists the 
actual values used (made up from stan- 
dard E24 series resistors). Column 5 
lists the actual values of R r obtained 


and column 6 lists the actual values of 
attenuation obtained using these resistor 
values. 

Resistor values for values of Rt other 
than 100 k can be obtained simply by 
scaling the resistor values given. For 
example, for a 50 k control the values 
should ail be halved, for a 10 k control 
they should be divided by 10 and so 
on. 

One final point to note is that the 
switch contacts should be of the make- 
before-break type to avoid switching 
clicks as the control is operated. M 


real load 
resistors 

When measuring and comparing the 
output powers of audio amplifiers 
(especially at the high end of the audio 
spectrum) it is useful to have available 
a ‘real’ load resistor, i.e. one which is a 
pure resistance with no parasitic induct- 
ance or capacitance. Carbon film re- 
sistors have a low self-inductance, but 
unfortunately are not commonly avail- 
able in the high power ratings required 
for amplifier testing. The highest rating 
normally available in a carbon film 
resistor is 2 watts, so a load resistor for 
testing a 100 W amplifier would need to 
be made up of 50 such resistors in 
series/parallel combinations! 

Wirewound resistors are available with 
high power ratings, but unfortunately 
such resistors are rarely wound so as 
to minimise self-inductance. A typical 
high-power wirewound resistor consists 
of a single layer of resistance wire 
wound helically on a cylindrical ceramic 
tube. This type of resistor has quite a 
high self-inductance, but since the usual 
applications of high-power wirewound 
resistors are DC or low-frequency AC 
this is not important. 



For use as an amplifier load resistor 
some means must be found of reducing 
the inductance of a wirewound resistor. 
This can be achieved by providing the 
resistor with a centre tap and connec- 
ting it as shown in figure 1. Current 
flows in opposite directions in each 
half of the resistor, so the magnetic 
fields produced in each half (and 
hence the self-inductances) tend to 
cancel out. If the original resistor has 
a value R then the connection shown 
has a resistance R/4 since it consists of 
two R/2 sections in parallel. 

Resistors already provided with taps, 
such as television H.T. dropper resistors, 
are suitable for this application. 
Presettable resistors may also be used. 
These consist of an exposed wire 
element wound on a ceramic former, 
and are provided with contact clips that 
may be fixed anywhere along the length 
of the element. 1 kW electric fire 
(heating) elements (which have a resist- 
ance of around 60 Cl) may also be used. 
In order to obtain a load resistor of the 
desired resistance and wattage rating, 
several wirewound resistors may be 
connected in series/parallel combi- 
nations in the normal way, provided 
each one is first connected as shown to 
minimise its inductance. M 




4-14 — elektor april 1978 


compander 



The dynamic range of an audio signal is 
the ratio, expressed in decibels (dB), be- 
tween the largest and smallest usable 
signal levels, i.e. between the loudest 
and softest sounds. ‘Live’ sound, from 
the softest whisper to the clatter of a 
pneumatic drill, can have a dynamic 
range in excess of 100 dB. However, it is 
not possible to capture such a large 
dynamic range in a recording, since the 
largest signal that can be recorded is lim- 
ited by saturation of the recording me- 
dium, and the smallest usable signal is 
limited by the recording medium’s own 
inherent noise, e.g. tape noise or record 
surface noise. The ratio between these 
two, i.e. the dynamic range, is only 
about 60 dB for the best disc recordings, 
and considerably less (around 45 dB) 
for cassette recordings. 

One way round the problem is to com- 
press the dynamic range of the original 
programme material before recording it, 
i.e. to pass the signal through a system 
whose gain reduces progressively as the 
signal level increases. Thus a 2 dB 
change in signal level at the input could 
be compressed, for example, into a 
1 dB change in level at the compressor 
output. To recreate the original dy- 
namic range the compressed recording 
is ‘expanded’ by replaying it through a 
system having the reciprocal transfer 
characteristic of the compressor, e.g. a 
circuit which gives a 2 dB change in 
output for a 1 dB change in input level. 

Disc to cassette 

The dynamic range of material recorded 
on disc is compressed into the 60 dB dy- 
namic range of this recording medium, 
but discs are normally played back 
without expansion since a dynamic 
range of 60 dB is considered adequate 
for domestic listening. Quite recently, 
DBX have introduced a disc com- 
pression/expansion system, but any ex- 
pander system will, of course, add to 
the cost of disc reproduction equip- 
ment. 

Transcribing discs onto cassette tape 
using a deck already equipped with a 
noise reduction system (Dolby, ANRS) 
is no problem. However, there are many 
inexpensive cassette decks on the mar- 
ket that are not equipped with a noise 


'Compander' is a portmanteau 
word derived from 'compressor' 
and 'expander' and describes a 
device designed to increase the 
dynamic range and/or improve the 
signal-to-noise ratio of an audio 
transmission, or recording and re- 
production chain. Until recently, 
companders were fairly compli- 
cated circuits. Now, however, they 
are available in the form of inte- 
grated circuits, one of which, the 
Exar XR 2216, is discussed in this 
article. This 1C can be used in a 
variety of applications such as 
amateur radio, PA systems, tran- 
scription of recorded material 
from disc to tape etc. 


Table 1. Electrical specification of the 
XR 2216. 


reduction system, and the results ob- 
tained when discs are recorded using 
such a machine are likely to be disap- 
pointing, since the dynamic range is in- 
adequate. Recordings made taking care 
not to overload the tape will have ex- 
cessive background noise on quiet pass- 
ages, while recordings made to give a 
reasonable noise level on quiet passages 
will exhibit distortion due to over- 
loading on loud passages. 

Normally, the only way to improve 
matters is to control the dynamic range 
of the programme material manually 
during recording, by ‘riding’ the re- 
cording level control. This can be very 
tedious if long passages are to be re- 
corded, so a simple compander would 
be a useful addition to an inexpensive 
cassette machine. 

The XR 2216 

Until recently companders were fairly 
complex circuits, but fortunately a com- 
plete compander system is now available 
in the form of an integrated circuit from 
Exar — the XR 2216. 

The equivalent circuit and functional 
block diagram of this 1C are shown in 
figure 1. The device contains an AC/DC 
converter which converts the AC signal 
fed to it into a proportional DC control 
voltage, a voltage-controlled impedance 
converter (which functions as a voltage- 
controlled attenuator) and a high-gain 
operational amplifier. 

Figure 2 shows the external components 
and circuit connections necessary to 
make the XR2216 function as an ex- 
pander. The input signal (from the tape 
deck, for example) is applied to pin 7, 
the input of the AC/DC converter, the 
output of which controls the transcon- 
ductance of the impedance converter. 
The input signal is also fed to the im- 
pedance converter, the output of which 
is thus proportional to the product of 
the input signal and its average value 
from the AC/D C converter, i.e. the 
transfer function of the expander is a 
square law. The impedance converter 
output is fed to the operational ampli- 
fier by linking pins 11 and 16, and the 
expanded output signal is taken from 
pin 2. 

By re-arranging the circuit slightly it can 


compander 


elektor april 1978 — 4-15 


ELECTRICAL CHARACTERISTICS: V c = +12 V, Ta = 25 C 


COMPANDER 


PARAMETERS 


Power Supply Voltage 


Nominal Power Supply Voltage 


Power Supply Current, No Signal Input 


Gain Change Over Frequency Tolerance 


Distortion Measured at —4 dB* 
Input Level at 1 KHz 


Attack Time Measured at —10 dB 
Input Level 


Decay Time Measured at —10 dB 
Input Level 


Transfer Characteristics** 

Compander Output With Input Levels of: 

- 4dB 

- 8dB 
-10 dB 

— 14 dB (reference) 

-24 dB 
-34 dB 
-44 dB 
-54 dB 
-64 dB 


MIN. TYP. 


6 


12 



MAX. 

UNITS 

20 

VDC 

18 

VDC 

3 


+ 1 

dB 


% THD 

5 

ms 

5 

ms 


CONDITIONS 


3.5 

+6 

7.5 

dB 

-0.5 

+2 

3.5 

dB 

-1.5 

0 

+ 1.5 

dB 


-4 


dB 

-15.5 

-14 

-12.5 

dB 

-25.5 

-24 

-22.5 

dB 

-36.5 

-34 

-32.5 

dB 

-49 

-44 

-42.5 

dB 

-59 

-54 

-52.5 

dB 



To 10% of Final Value 



COMPRESSOR 

PARAMETERS 


Input Impedance 


; Output Impedance 


Output Signal Level for —10 dB 
Input at 1 KHz 


Output Voltage Swing 


Output Noise, Input AC Grounded 


Compressor Transfer Characteristics** 

Compressor Output With Input Levels of: 

- 4dB 

- 8dB 
-10 dB 

—14 dB (reference) 

-24 dB 
-34 dB 
-44 dB 
-54 dB 
-64 dB 


MIN. TYP. MAX 


50 


X. UNITS 


k ohm 




CONDITIONS 



EXPANDER 


PARAMETERS 


Input Impedance 


Output Impedance 


Output Signal Level for —10 dB 


Output Voltage Swing 


Output Noise Input AC Grounded 


Expander Transfer Characteristics** 

Expander Input Levels Required for Output of: 
+ 6dB 
+ 2dB 
0 dB 

— 4 dB (reference) 

-14 dB 
-24 dB 
-34 dB 
-44 dB 
-55 dB 


MIN. TYP. MAX. 


50 


kX. UNITS 


k ohm 










CONDITIONS 



Notes: * 0 dB = 0.775 Vrms (1 mW across 600 ohm load) ** Recommended transfer characteristics. 
































































4-16 — elektor april 1978 


compander 




be made to function as a compressor, 
the circuit of which is shown in figure 3. 
In this case the input signal is fed to the 
input of the impedance converter 
(pin 10) and from the output of this 
stage to the input of the operational 
amplifier, the output (to the tape deck) 
again being taken from pin 2. A portion 
of the output signal is fed to the input 
of the AC/DC converter, (by bilking 
pins 2 and 7), the output of which again 
controls the transconductance of the 
impedance converter. In this case the 
output is thus proportional to the 
square root of the input signal, i.e. the 
transfer function is the reciprocal of the 
expander circuit’s. 

The attack and decay times of the cir- 
cuit are equal and are determined by a 
filter consisting of an external resistor 
and capacitor (P1/C3 or P3/C8). It is 


Figure 1. Internal circuit and functional block 
diagram of the XR 2216 compander 1C. 

Figure 2, Connections and external compo- 
nents required for operation of the 1C as an 
expander. 

Figure 3. Connection of the XR2216 as a 
compressor. 

Figure 4. Typical performance curves of the 
XR 2216: 

4a. Compressor output error versus signal am- 
plitude. 

4b. Expander input error versus output signal 
amplitude. 

4c. Compander tracking error versus input sig- 
nal amplitude. 


important that the attack time should 
not be too long, otherwise the response 
of the circuits to transients may be too 
slow to prevent overload. 

On the other hand, if the decay time is 
too short then ripple may appear on the 
output of the AC/DC converter at low 
input frequencies, thus leading to modu- 
lation of the output signal and third- 
harmonic distortion. This is not a prob- 
lem in a compander system, since the 
distortions produced in the compression 
and expansion processes tend to cancel 
out. However, if the circuit is used 
simply as a compressor or as an ex- 
pander then distortion at low fre- 
quencies is a major problem. 

Two preset adjustments are provided in 
the circuits of figures 2 and 3. PI and 
P3 set the reference level of the circuit, 
which determines the actual input volt- 


age range over which the compression/ 
expansion takes place. P2 and P4 set the 
low level tracking, which ensures that 
the compression and expansion charac- 
teristics match, thus ensuring (amongst 
other things) minimum distortion. 


Performance data 

The specifications of the XR2216 are 
given in table 1, and typical perform- 
ance curves in figure 4. It can be seen 
that, in the compressor mode the circuit 
provides a 2 : 1 compression ratio, e.g. a 
60 dB dynamic range can be compressed 
into 30 dB. In the expansion mode, not 
surprisingly, an expansion ratio of 1 : 2 
is obtained, thus restoring the original 
dynamic range. 

From table 1 it can be seen that distor- 







compander 


stereo pan pot 


elektor april 1978 — 4-17 


4a 


V CC -12V. f~1kHz 


CL 

o 

CL 

CL 

Ui 

I- 

D 


o?0 

CL 

o 


5 

o 

o 


B 

i 

1 


B 

1 

* 

n 


1 

i 

B 

a 

SB 

i 

s 

1 


B 

a 

a 

i 

1! 

a 


S 

i 

I 


-70 -60 -50 -40 -30 -20 -10 

COMPRESSOR INPUT SIGNAL 
dBm600fi 


4b 


Vcc= 12 V, f - 1 kHz 


oc 

o 

cc 


i 

1 


1 

n 

A * + 



i / 

60° C / 

ll 

■ 

a 

■ 

■ 

| 

| 

i 



B 

8 

i 

i 

m 


i 

1 




1 

1 

1 


EXPANDER OUTPUT SIGNAL 
dB 



+25° C 


tion at 1 kHz is typically 3%. Whilst this 
may not seem particularly low it is cer- 
tainly comparable with the distortion 
level of an inexpensive cassette recorder, 
and is more than adequate for non-hi-fi 
purposes. 

It should be noted that the values of C6 
and CIO shown on the circuit diagram 
were calculated for circuits having an in- 
put impedance of around 50 k. If the 
equipment being used with the com- 
pander has a lower input impedance, the 
values of these capacitors can be calcu- 
lated by using the formula shown below: 


c -tr 


(|UF, kfi) 



R (in kf2) being equal to the input im- 
pedance in parallel with R1 (or R2), and 
C being C6 or CIO (in /iF). H 


When making a sound recording using 
multi-microphone techniques the signal 
picked up by each microphone can be 
correctly positioned in the stereo sound 
stage by ‘panning’. For example, the sig- 
nal from a centrally placed microphone 
would be fed equally to both left- and 
right channels, a signal from a micro- 
phone located at the left of the sound 
stage would be fed only to the left chan- 
nel and a signal from a microphone lo- 
cated at the right would be fed only to 
the right channel. Signals from micro- 
phones located between these positions 
would be fed to the left- and right chan- 
nels in the appropriate proportions. 

A circuit which allows the position of 
the sound image from a particular 
microphone to be positioned is known 
as a panoramic potentiometer or pan 
pot and usually consists of a ganged log- 
antilog potentiometer. The input signal 
fed to both halves of the poten- 
tiometer and the left- and right outputs 
are taken from the wipers. Turning the 
potentiometer to the right increases the 
right channel level and decreases the left 
channel level, and vice versa. 

Operation of a pan pot must not vary 
the total signal level, i.e. if the output 
level from the left or right channel with 


the pan pot in its extreme left- or right 
position (other channel muted) is taken 
as 0 dB, then the signal level from each 
channel with the pan pot central must 
be -3 dB to keep the total signal con- 
stant. 

Figure 1 shows the circuit of a pan pot 
which uses only a single, linear poten- 
tiometer. The input signal from, say, a 
microphone preamplifier is split into 
two channels. The resistor and poten- 
tiometer values are chosen such that, 
with PI in the extreme left position 
(wiper towards R 1 ’) the gain of the left 
channel is 1.066 whilst the right input 
signal is shorted to ground via the wiper 
of PI . With PI in the extreme righ pos- 
ition the reverse is true. With the wiper 
of PI in its centre position the gain of 
both channels is 0.746, which is ap- 
proximately 3 dB down on the gain in 
the extreme positions. 

National Semiconductor appl. M 


R2 Cl 

rQQrCEHKH 

i/i 


o 

CH 



’ 

K 



i R2‘ Cl* 

^cehH 

ip 


IC1.IC1-LM387 or equ. 










4-18 — elektor april 1978 


723 as a constant current source 



Figure 1 shows a simplified internal cir- 
cuit of the /UA723, equivalents for 
which are the LM723 and TBA281. It 
contains a temperature-compensated 
voltage reference, a differential ampli- 
fier, driver and output transistors and 
a current sense transistor for current 
limiting purposes. A temperature- 
compensated reference voltage of 
7.15 V +/— 5% is available at pin 4 
(metal can version) or pin 6 (DIL pack- 
age version). Familiarity with this 
internal circuit will aid in understanding 
the operation of the 723 as a constant- 
current source, which is shown in fig- 
ure 2. 

The differential amplifier is connected 
as a voltage-follower, with the output 
V 0 fed directly back to the inverting 
input. A potential divider, R2/R3, 
connected across the reference voltage 
output, feeds a voltage of about 2.2 V 
to the non-inverting input. Since the 
differential amplifier is connected as a 
voltage follower, 2.2 V appears at out- 
put V Q . This causes a constant current 


to flow through Rl. Since this current 
flows from the positive supply rail into 
the V c pin, it must also flow through 
the external load Rl. This current is 
constant, irrespective of the value of 
RL, within certain limits. The maximum 
value of Rl is given by: 

rl- V - (a v, a). 

Although the maximum output current 
capability of the 723 is 150 mA, care 
must also be taken not to exceed the 
800 mW maximum dissipation of the 
IC. Maximum dissipation occurs when 
Rl is zero, since almost all the supply 
voltage is then dropped across the out- 
put transistor of the 1C. The dissipation 
is given by: 

P = (Ub - 2.2) x I (W, V, A). 

Rearranging this equation and substitut- 
ing 0.8 W as the maximum dissipation, 
the maximum current that can safely 
be supplied (into a short-circuit) is 


Imax ~ 2~2 2 (A, W, V). 

With a 10 V supply this is approxi- 
mately 100 mA, and with the maximum 
supply (37 V) it will be approximately 
23 mA. 

The 723 may be provided with a ther- 
mal shutdown facility to protect against 
overheating. This is achieved by using 
the current limit transistor in the IC as a 
temperature sensor. At 30°C the base- 
emitter ‘knee’ voltage of this transistor 
is about 0.65 V, but at 120°C it has 
fallen to about 0.5 V. Resistors R4 and 
R5 (shown dotted) apply approximately 
0.5 V to the base of this transistor (note 
also the dotted connection to the C s 
terminal). This is normally less than the 
base-emitter knee voltage and is insuf- 
ficient to turn on the transistor, but at 
120°C, when the knee voltage has 
dropped to 0.5 V, the transistor will 
start to turn on. This will reduce the 
base drive to the IC’s output stage, 
decreasing the output current and hence 
the dissipation. 

If a larger output current is required 
than can be provided by the ;uA 723, an 
external power transistor may be added, 
as shown in figures 3 and 4. If an NPN 
transistor is used then it is simply con- 
nected as an extension of the emitter- 
followers in the IC’s own output stage: 
base to V 0 , emitter to the inverting 
input of the differential amplifier. How- 
ever, if a PNP transistor is used a slight 



The nA 723 precision voltage 
regulator IC is well known for its 
versatility, good line and load 
regulation, and low temperature 
coefficient. In addition to its 
many uses as a voltage regulator, 
it can also be used as a precision 
current regulator (constant 
current source). 





723 as a constant current source 


coming soon 


elektor april 1978 — 4-19 



comm 

=soo: 


ing 

oon 



-Mv* v+v = Vo “ 
I TBA281 
I mA723 




V — comp | 


i 

* 

R, n 

Cl mim 


jUJOp 





digital reverb unit 



Figure 1. Simplified internal circuit of the 
723 1C regulator. Numbers in parentheses are 
pinout of the DIL package version; others, 
pinout of the TO-metal can version. 

Figure 2. The 723 used as a constant current 


Figures 3 and 4. If a larger output current is 
required than can be provided by the 723 
alone, an external NPN or PNP power transis- 
tor may be added. 


rearrangement of the circuit is necess- 
ary: V 0 and the inverting input are 
linked, and the base of the transistor is 
connected to V c , the collector of the 
IC’s output transistor. The equation 
previously given for calculating the out- 
put current also holds for these two 
circuits. 

The thermal shutdown facility may 
also be added to these circuits, but it 
should be emphasised that it will 
protect only the IC, not the external 
transistor. As the dissipation in the 
external transistor may be quite high it 
is essential to provide it with a substan- 
tial heatsink. For example, with a 37 V 
supply and a current of 1 A, the short- 
circuit dissipation in the external tran- 
sistor will be about 35 W! M 


^s 0.0 00 


frequency counters 


colour modulator 
mini shortwave receiver 
universal logic tester 
percolator switch 
traffic lights 
etcetera etc. 






4-20 — elektor april 1978 


cassette interface 


cassette- 

interface 


In contrast to some microprocessors, 
the SC/MP possesses serial input and 
output ports, so that all parallel-serial 
and serial-parallel conversion can be 
effected by software control. The 
necessary software routines, which 
ensure that data is read serially into and 
out of memory, are already contained in 
Elbug, the monitor software for the 
Elektor SC/MP system, and were dis- 
cussed in the previous article in the 
series. This month we concentrate upon 
the hardware needed to convert the 
serial digital information into an ana- 
logue signal suitable for storage on tape. 
There are several different systems for 
encoding digital information into a form 
suitable for recording. The most com- 
mon of these, and the one employed in 
this circuit, is the CUTS format. CUTS 
is the acronym for Computer Users 
Tape System, and is sometimes also 
called the Kansas City Standard, in 
addition to specifying the number of 
control bits and the transmission speed 
(300 Baud), it defines that a logic *1* be 
encoded as eight cycles of a 2,400 Hz 
audio tone, whilst a logic ‘0’ be re- 
corded as four cycles of 1 ,200 Hz. 
These frequencies were deliberately 
chosen as being suitable for use with all 
types of tape recorder. 


Cassette encoder/decoder 

The encoder/decoder consists of an FSK 
modulator (FSK = Frequency Shift 
Keying) and an FSK demodulator. 
When a logic ‘1’ is present at the input, 
the modulator outputs a sinewave with 
a frequency of 2,400 Hz. If the input 
signal is at logic ‘0’, the frequency of 
the output signal shifts to 1 ,200 Hz. 
The modulator output is fed direct to 
the cassette recorder input. 

When the recorder is playing back an 
encoded programme, the output of the 
recorder is fed to the FSK demodulator, 
which will output a logic ‘1’ if the fre- 
quency of the input signal is 2,400 Hz, 
or a logic ‘(5’ if the frequency is 
1,200 Hz. The demodulator output is 
connected to the serial input of the 
SC/MP, and a software routine ensures 
that the serial stream of data is con- 


Low cost, high transmission rate 
and complete reliability were the 
design requirements for the 
cassette interface described in the 
following article. The interface 
makes very few demands on the 
sound quality of the recorder, and 
can transfer data at a rate of up to 
1200 Baud. 



verted back into parallel form. 

The speed of transmission is controlled 
by the software, and not by the inter- 
face hardware. Several different trans- 
mission rates are possible (see Elektor 
35) and without any adjustments the 
interface can be used at speeds up to 
1200 Baud. 

Transmission rates higher than this are 
not possible with the above frequencies, 
however it is debatable whether they are 
in fact desirable, since with a speed of 
1,200 Baud it would only take approx. 
10 minutes to record a programme of 
64 k bytes (i.e. the entire memory ca- 
pacity of the SC/MP). 


FSK modulator 

The FSK modulator uses a well-known 
IC function generator, the XR-2206 (see 
figure 2). The IC is powered by the two 
supply voltages already present in the 
SC/MP system, i.e. + 5 V and 12 V. 
Transistor T1 is included so that the IC 
can be driven by TTL logic levels. 

The input and output signals of the 
modulator are shown in figures 5a and 
5b, respectively. The amplitude of the 
output signal can be varied by means of 
P3, and hence adjusted to suit the input 
sensitivity of the particular tape re- 
corder being used. The frequency of the 
output signal can be set to 1,200 Hz and 
2,400 Hz by means of PI and P2 re- 
spectively. This can be done either by 
using a frequency meter, or if that is not 
possible, by utilising the SC/MP clock 
generator. 


Tuning the modulator 

Since the SC/MP has an internal crystal 
clock generator, it is possible, with the 
aid of a short programme, to get it to 
generate signals whose frequency is re- 
markably constant. Table 1 shows the 
listing for such a programme. Once it 
has been loaded into memory and run, a 
squarewave signal with a frequency of 
either 1,200 Hz or 2,400 Hz is available 
at Flag 0 (pin 14C of the connector 
bus). The actual frequency of the signal 


cassette-interface 


elektor april 1978 — 4-21 


Figure 1. Block diagram of the cassette 
interface. 

Figure 2. Circuit diagram of the FSK modu- 
lator. 

Figure 3. Simple circuit for tuning the modu- 
lator. 


Table 1. Listing of the programme to tune the 
modulator. 



depends upon the ‘number’ written into 
•XX’ and ‘YY’ (see table 1). 

These signals can then be used to tune 
the modulator as follows: 

• Flag 0 and the modulator output are 
connected as shown in figure 3. A 
high impedance earphone or a tape 
recorder with input level meter is 
then connected to the output of the 
above circuit. 

• The programme for the 1 ,200 Hz 
tone is started, and a logic ‘0’ is 
presented to the modulator input 
(potentiometer P3 is set for maxi- 
mum amplitude). Several different 
frequencies should now be audible in 


the earphone, namely the 1,200 Hz 
signal produced by the SC/MP, the 
tone produced by the modulator and 
the difference- or beat signal. 

• PI is adjusted until the beat signal 
frequency is reduced to a minimum, 
when the frequency of the modu- 
lator output should be virtually the 
same as that of the programme signal. 
If using a VU-meter then PI is 
adjusted until the needle ceases to 
jitter. 

The procedure for the 2,400 Hz signal is 
exactly the same, except that the 
SC/MP is programmed to produce a 


2,400 Hz signal and a logic ‘1 ’ is applied 
to the modulator input. 


FSK demodulator 

The circuit for the demodulator is con- 
siderably more complicated than that 
for the modulator. The complete circuit 
diagram of the FSK demodulator is 
shown in figure 4. Its operation differs 
from the conventional method of FSK 
demodulation (PLL), but the series of 
voltage waveforms shown in figure 5 
should help to simplify its explanation. 
The FSK signal (figure 5b) is fed to the 














cassette interface 


4-22 — elektor april 1978 


5V 



4 «B i 09 l)" 

==> IC1 SE IC2 “ IC3 


5V 

o 


1 


47»i 

6V 


IC2 
,0 °" (?) 


D3 . . D6 = 1N4148 
A1 ... A3 = IC2 = CA 3060 
N1 . . . N6 = IC3 = 4049 
MMV 1 , MMV 2 = IC4 = 74123 


o 


input of de demodulator, where it is 
clipped symmetrically by the limiter 
amplifier A1 (figure 5c) before being 
fed to a Schmitt trigger (Nl, N2). The 
output of the trigger and its inverted 
form are each fed to a differentiating 
network (C9, R18 and CIO, R19). The 
result is that a signal which has twice 
the frequency of the original input 
signal of the Schmitt-trigger is present at 
the collectors of T2 and T3 (see figure 
5d). This signal is then used to gate a 
monostable multivibrator (Ml). The 
output of the monostable is a train of 
pulses of constant width (figure 5e). 


These pulses are fed to an integrating 
network (R22, Cl 3), producing the sig- 
nal shown in figure 5f. 

The voltage across capacitor Cl 3 (fig- 
ure 5f) can be used to monitor the 
frequency of the input signal, since the 
increase in charge per unit of time is 
twice as great in the case of the higher 
frequency (2,400 Hz) than in the case 
of the lower frequency (1,200 Hz). In 
order to convert this small difference in 
the DC component of the voltage across 
C13 into a digital signal, a sample-and- 
hold circuit (A2, T4 and Cl 4) is used as 
a memory. A sample-pulse is derived 


from the output of Ml by means of a 
second monostable (M2). During this 
sample-pulse (figure 5g) the sample-and- 
hold circuit samples and then stores the 
instantaneous value of the voltage across 
Cl 3. If the frequency of the demodu- 
lator input signal is constant, then the 
voltage at the output of the sample-and- 
hold circuit (= emitter of T4) will also 
be virtually constant. If the frequency 
of the input signal varies, then at the 
moment of sampling, the instantaneous 
value of the voltage across Cl 3, and 
hence the output voltage of the sample- 
and-hold (figure 5h) will also vary. 























cassette interface 


Figure 4. The circuit diagram of the FSK de- 
modulator. 

Figure 5. Diagram of the various voltage wave- 
forms at different points in the demodulator 
circuit. 

Table 2. Programme for tuning the demodu- 
lator. 


elektor april 1978 — 4-23 



Although this output voltage is in fact a 
digital signal, its amplitude and logic 
voltage swing are not very large, and 
hence a comparator (A3 and N5, N6) is 
required to bring the signal up to TTL- 
logic levels. The threshold voltage of 
this comparator in fact represents the 
only adjustment point in the entire 
demodulator, and once again a simple 
programme will prove useful. 


Tuning the demodulator 

If the demodulator is fed a ‘symmetrical’ 



START 

= 0C00 


0C00 

C404 

LDI 04 


0C02 

C833 

ST 33 

; count 2400 Hz 

0C04 

C401 

LDI 01 

; periods 

0C06 

07 

CAS 

; Set F lag 01 

0C07 

08 

NOP 


0C08 

C41E 

LDI IE 


0C0A 

8F00 

DLY 00 

; Delay 

0C0C 

C400 

LDI 00 


0C0E 

07 

CAS 

; Reset Flag 0 

0COF 

C402 

LDI 02 


0C11 

8F00 

DLY 00 

; Delay 

0C13 

B822 

DLD 

; count periods 

0C15 

9804 

JZ 


0C17 

F000 

ADD 00 

; Timing correction 0 

0C19 

90E9 

JMP 


0C1B 

C402 

LDI 02 


0C1D 

C819 

ST 19 

; count 1200 Hz 

0C1F 

C401 

LDI 01 

; periods 

0C21 

07 

CAS 

; Set Flag 0 

0C22 

08 

NOP 


0C23 

C452 

LDI 52 


0C25 

8F00 

DLY 00 

; Delay 

0C27 

C400 

LDI 00 


0C29 

07 

CAS 

; Reset Flag 0 

0C2A 

C436 

LDI 36 


0C2C 

8F00 

DLY 00 

; Delay 

0C2E 

B808 

DLD 


0C30 

98CE 

JZ 


0C32 

F000 

ADD 00 


0C34 

90E9 

JMP 


0C36 

00 

00 

; 2400 Hz counter — 1 

0C37 

00 

00 

; 1 200 Hz counter — 1 




4-24 — elektor april 1978 


cassette interface 


19905 


■V 

i 


Parts list to figures 6 and 7 

Resistors: 

R1 = 1 k 
R2 = 4k7 

R3,R6,R18,R19,R23 = 12 k 
R4 = 5k6 

R5,R1 1,R1 3,R20,R21,R22, 
R26.R27 = 22 k 
R7,R1 2,R24,R34 = 220 n 
R8.R9 = 10 k 
RIO = 18 k 
R1 4 = 15 k 
R15.R28 = 100 k 
R1 6 = 1 M 
R17 = 1 k5 
R25 = 27 k 
R29 = 220 k 
R30 = 470 k 
R31 = 2k7 
R32 = 470 ft 
R33 = 3M9 
PI = 10 k 
P2 = 5 k 
P3 = 25 k 
P4 = 1 k 










c v > 
c > 

C _ 3 


a 

B 

-3 

u 

u 

A 








Q jnia | p L 

I® 


n<HK> q p 6 




S 


o o o o 

+ 1 O 




Capacitors: 

Cl = 27 n 
C2,C3 = 10 m/16 V 
C4 = 1 m/ 16 V 
C5 = 1 80 n 
C6 = 1m5/6 V 
C7 = 100 p 
C8 = 56 n 
C9,C10,C1 5 = 1 n 
Cl 1, Cl 3= 10 n 
Cl 2 = 150 p 
C14 = 1 n5 
Cl 6 = 4n7 
C17 = 47 m/6 V 
C18,C19 = 100 n 
C20 = 3n3 

Semiconductors: 

D1 D6= 1 N4148 

T1 = BC 557, TUP 

T2,T3 = BC 547 TUN 

T4 = BC517 

IC1 = XR-2206 

IC2 = CA 3060 

IC3= 4049 (CD 4049, etc.) 

IC4= 74123 


Figure 6. Track pattern of the interface board 
(EPS 9905). 

Figure 7. Component layout of the interface 
board. 

Figure 8. Wiring diagram for the connections 
between the SC/MP system (bus board) and 
the cassette interface. 






cassette interface 


elektor april 1978 — 4-25 



input signal, i.e. a signal consisting of 
equal-lengthed portions of 2,400 Hz and 
1,200 Hz, then the output signal must 
also be symmetrical. Table 2 lists a pro- 
gramme which will generate a sym- 
metrical input signal for the demodu- 
lator. This signal is available at Flag 9 of 
the SC/MP and hence the demodulator 
input should be connected to this point 
(connector pin 14C). 

The output signal is adjusted by means 
of P4. Since a symmetrical signal which 
swings between supply and earth has an 
average value which is equal to half 
supply, the demodulator output should 
be connected to a DC voltmeter and P4 
adjusted until a reading of 2.5 V is ob- 
tained. That completes the adjustment 
procedure for the demodulator. 


Printed Circuit Board 

A printed circuit board was designed to 
accommodate both the modulator- and 
demodulator circuit. Figures 6 and 7 
show the track layout and component 
overlay of the board. Once the compo- 
nents have been mounted and both 
circuits correctly adjusted, the board 
can be connected to the SC/MP system 
as shown in figure 8. H 


is 



car rip-off protection 


4-26 - elektor april 1978 


car rip-off 
protection 


The complete circuit of the alarm is | 
shown in figure 1. N1 to N4 form a 
5-input OR gate, but the number of 
inputs can easily be increased by adding 
extra gates. When the car is unoccupied 
(and the ignition is switched off) R1 1 
holds the inputs of N5 low, so the out- 
put is high. The inputs of N1 to N4 are 
held low via the filaments of the lamps, 
etc. that are being protected. The out- 
put of N4 is thus low, the output of N6 
is high, T1 is turned on, T2 is turned off 
and relay Re.l is de-energised. 

In the event of an accessory being 
disconnected by a thief (for example, 
the lamp connected to input El), then 
the appropriate input to the OR gate 
will be pulled high by the 10 k input 
resistor. The output of N4 then goes 
high, the output of N6 goes low, T1 is 
turned off and T2 is turned on, ener- 
gising Re.l and sounding the car horn. 
When the ignition switch is closed the 
output of N5 is low, which holds the 
output of N6 permanently high, thus 
disabling the alarm. This prevents the 
alarm from sounding when one of the 
accessories is switched on. Of course, 
the alarm will still sound if an accessory 
is switched on whilst the ignition is 
switched off. This prevents spotlamps 
and foglamps from accidentally being 
left on whilst the car is unoccupied. 
Alternatively, these accessories can be 
wired via the ignition switch so that this 
cannot occur. 

An additional bonus is that the alarm 
will also sound in the event of a lamp 
filament failure. However, since a 


Thefts from cars of valuable 
accessories such as spotlamps and 
foglamps are on the increase. 
Equipped with a spanner, and 
given a few minutes undisturbed, 
the enterprising felon can 
frequently make a haul worth over 
£100. The inexpensive alarm 
circuit described in this article will 
protect these valuable items, and 
can also be used to prevent the 
theft of accessories from inside 
the car, for example radios and 
cassette players. 

W. Braun 


Figure 1. Complete circuit of the car rip-off 
theft alarm. 

Figure 2. Lamps connected in pairs must be 
isolated from one another, otherwise the 
alarm will not give complete protection. This 
can be done with a pair of diodes, as shown in 
figure 2a, or by double-pole switching, as 
shown in figure 2b. 


replacement lamp may not always be 
available, a secret ‘cancel’ switch (SI) is 
required . . . 

Accessories inside the car, such as radios 
and cassette players, may also be pro- 
tected by connecting a wire from one of 
the alarm inputs to the earthed case of 
the equipment. When the thief cuts this 
wire to remove the equipment then the 
alarm will sound. Of course, this facility 
should only be regarded as a backup to 
an alarm that prevents a thief entering 
the car in the first place ! 

Where lamps are wired in pairs the alarm 
will, of course, not sound until both 
have been disconnected. To overcome 
this disadvantage a diode of suitable 
current rating can be wired in series 
with each lamp, as shown in figure 2a. 
Since there is a 0.7 V voltage drop across 
a diode, a better idea would be to use a 
double-pole switch for the set of lamps 
being protected, as shown in figure 2b. K 




vocoders 


elektor april 1978 — 4-27 



A vocoder (VOice CODER) is an in- 
strument designed to analyse and 
electronically recreate the sound of the 
human voice. Although vocoders are in 
fact a far from recent invention, and 
have been used for a number of years in 
such fields as telecommunications and 
data processing, it is only within the last 
couple of years that a serious attempt 
has been made to exploit their enormous 
potential for musical and sound effect 
applications. 

History 

The term ‘vocoder’ was first coined in 
1936 by an American called Homer' 
Dudley, who invented a machine to 
compress the bandwidth of speech for 
transmission purposes. There was also a 
certain amount of interest in vocoders 
in Germany during the thirties. This 
interest was stimulated by the realisation 
that they had an obvious military po- 
tential — the encoding of secret mess- 
ages. 

By the middle of the sixties Siemens 
possessed a vocoder which was occasion- 
ally used for recordings. Similarly the 
BBC Radiophonic Workshop, and a 
number of other experimental studios 
used vocoders for special effects on 
records, radio and television. However 
all these early prototypes suffered from 
the drawback of being extremely large 
and unwieldy, and as such were quite 
unsuited for other than specialised 
applications. 

The real breakthrough came in 1975 
with the appearance of a vocoder which, 
by virtue of its compact and ergonomical 
design, was suitable for use in a conven- 
tional studio situation where it could be 
interfaced with other equipment, thus 
allowing its full potential to be realised. 
This was the EMS (Electronic Music 
Studios) Vocoder (see photo 4) 
developed by Tim Orr, a self-contained 
portable instrument that can not only 
synthesise speech at constant and varying 
pitch, but by using a second non-speech 
input signal can encode literally any 
recorded sound with any speech sound. 
The machine can thus produce the 
effect of ‘talking’ musical instruments. 
Since the EMS Vocoder, Sennheiser 
have capitalised upon their experience 


An orchestra suddenly begins to 
recite a passage of Shakespeare, an 
electric guitar reads the news, the 
voice of a talker unexpectedly 
changes sex, a single voice sounds 
like a chorus — these are just a few 
of the amazing effects which can 
be obtained with a new electronic 
instrument — the vocoder. 

This article explains the ins and 
outs of this fascinating new 
development in the field of 
electronic 'music'. 

C. Chapman 


The author and editor wish to thank Mr. Orr 
of EMS Ltd., Mr. Buder of Sennheiser and 
Mr. Funk of the Hamburg Radio Studio for 
their assistance in the preparation of this 
article. 


of using vocoders in the field of tele- 
communications, and with the assistance 
of Heinz Funk of the Hamburg Radio 
Studio have brought out the Sennheiser 
Sound Effect Vocoder VSM 201 (see 
photo 5). The latest development is a 
smaller version of the EMS Vocoder, 
called the EMS 2000 (see photo 6), 
which, by virtue of its size and extreme 
portability, is particularly suited for live 
work. 


Speech-synthesis and Vocoding 

As mentioned above, a fundamental 
feature of vocoders is their ability to 
analyse and electronically simulate the 
sound of speech. Thus before going on 
to examine the operating principles of a 
vocoder it is first necessary to take a 
look at the basic characteristics of 
human speech. 

Speech sounds 

At the moment it is virtually impossible 
to create a realistic replica of the human 
voice, since not only do speech sounds 
have a very irregular intensity, but they 
are also extremely rich in harmonics. 
Synthesised speech is always too ‘clean’, 
too free from natural imperfections. 
Speech itself is composed of two main 
component sounds: 

• Air from the lungs can be forced 
between the vocal chords situated in 
the windpipe, causing these chords to 
vibrate and a pulsating air-column to 
enter the mouth and nasal cavities. 
The fundamental frequency of the 
resultant note is determined by the 
length, thickness and tension of the 
vocal chords. Sounds produced in 
this fashion e.g. the vowels, are 
known as VOICED sounds. 

• Alternatively, if the air from the 
lungs is not forced through the 
vocal chords, but simply expelled 
through the mouth, then so-called 
UNVOICED sounds are produced, 
such as T or ‘h’. These are basically 
similar to the type of sounds which 
can be produced by a noise generator. 

In the case of both voiced and unvoiced 
sounds the shape of the mouth and 
nasal cavities determines the character 
or timbre of the sounds. Variation of 


4-28 — elektor april 1978 


vocoders 


speech 

articulation 


replacement 
signal 
input r\ 


speech 



cavity RESONANCES by movement of 
the tongue and lips controls the 
harmonic content of the voice and 
enables us to form separate vowels and 
consonants (see figures 2a and 2b). The 
lips play a particularly important role in 
sounds which are distinguished by their 
dynamic amplitude characteristics, such 
as the percussive attack transient of the 
‘p’ in ‘paper’. 

Thus the voice can be seen as a complex 
sound generating instrument, consisting 
of a frequency and amplitude-controlled 
oscillator (the vocal chords and lungs), a 


Figure 1. This simplified diagram shows the 
basic operating principle of all vocoders. The 
input speech signal is analysed to provide a 
set of data which is used to impose the pattern 
or articulation of the speech signal upon an 
external replacement signal input. The fact 
that the original speech sounds are encoded in 
the form of control voltages gives the vocoder 
its name (VOice CODER). 

Figure 2. A spectrum analysis of the sound 
of vowels and consonants spoken by a female 
(figure 2a) and a male (figure 2b) voice. The 
pitch is the same for all the vowels. The 
fundamental frequency of the male voice is 
approx. 140 Hz, whilst that of the female 
voice is roughly 280 Hz. The sound produced 
by the vibration of the vocal chords is 
extremely rich in harmonics. The variation in 
the dynamic amplitude characteristics of 
different vowels is the result of the different 
resonances formed by varying the position of 
the tongue, teeth and lips, and hence the shape 
of the nasal and mouth cavities. This process, 
which amounts to a sophisticated 'filtering' of 
the speech sounds, is just as important in the 
case of unvoiced sounds. This is evident from 
the differences in the spectra of the two 
sounds 'f' and 'sh'. 







vocoders 


elektor april 1978 — 4-29 


noise generator (the lungs) and a set of 
tone filters (the mouth and nasal 
cavities). 


Speech -synthesis 

Viewing the voice in this way naturally 
leads one to speculate whether it might 
be possible to synthesise speech, using 
techniques similar to those employed in 
a music synthesiser. The vocal chords 
could be replaced by an oscillator, the 
output waveform of which is suf- 
ficiently rich in higher harmonics to 


allow differentiated filtering, whilst a 
noise generator could be used to provide 
the unvoiced sounds. A switching circuit 
would cut back and forth between the 
above two sound sources depending 
upon which mode of voice was required. 

However problems begin to arise when 
one considers the type of filters that 
would be needed fora speech synthesiser 
of this type. Since the continual 
variation of both the static harmonic 
content and dynamic characteristics of 
the sound is crucial for the formulation 
of articulate speech, an equaliser-type 


filter system would be necessary to 
simulate all the nuances in the tonal 
character of human speech. What is 
more, the filter system would have to be 
voltage controlled if it were to have any 
chance of matching the rapid change in 
the harmonic content of speech. At this 
point it becomes clear that an analogue 
speech-synthesiser of this kind would 
require an enormous amount of 
hardware, for how does one generate 
the extremely complex pattern of 
voltages needed to control the filter 
bank? 

One possibility to simplify the process is 
a hybrid system, using a memory to 
store the control voltages. The quality 
of modern speech-synthesisers which 
use such a system is fairly good. Doubt- 
less many readers will have seen or 
heard of so-called ‘talking’ computers, 
which use synthetically-generated speech 
to express the results of their calcu- 
lations, and the ‘talking’ calculator 
shown in photo 1 proves that it does not 
require an enormous amount of 
hardware to synthesise speech digitally. 
Photo 2 shows that the digital speech- 
synthesiser consists of just two ICs 
mounted on a single board. The speech 
components are stored digitally in a 
ROM, where they can be scanned by a 
speech synthesiser micro-controller. A 
D/A converter in the micro-controller 
then generates the analogue speech 
components, from their digital equiv- 
alents. 

Vocoding 

Although storing the speech com- 
ponents digitally represents by far and 
away the simplest solution for systems 
designed to generate speech (assuming 
the desired vocabulary is not too large), 
this is not the case with vocoders, and 
here we come to the basic difference 
between vocoders and speech-syn- 
thesisers. 

A vocoder is basically designed to super- 
impose the pattern of spoken words 
onto a recorded non-speech signal (such 
as, music, the sound of wind, surf, etc.) 
so that the resultant effect is that of a 
talking orchestra, for instance. The 
articulation of the output signal is 
extremely good, being distinguished by 
remarkable clarity and distinctiveness. 
This quality of articulation, among 
other things, is what distinguishes the 
vocoder from other less sophisticated 
special effect devices such as the well- 
known WAWA pedal, or the more 
recent MOUTH BAG or MOUTH TUBE 
(see photo 3). 

The latter is basically a crude acoustic- 
mechanical vocoder. The signal from an 
electric guitar or similar source is fed to 
a powerful amplifier, which drives a 
loudspeaker situated in a closed box. 
The amplified sound from the guitar is 
then fed via a plastic tube to the mouth 
of the musician. Without using his vocal 
chords, but simply by altering the shape 
of his mouth cavity he can then articu- 
late the guitar signal, so that the guitar 




4-30 — elektor april 1978 


vocoders 



appears to be ‘talking’. This signal is 
picked up by a microphone in front of 
the musician’s mouth and fed through 
the PA system in the usual fashion. The 
sounds produced by the mouth tube are 
essentially similar to those produced by 
a vocoder. 

However, not only is the mouth tube 
fairly limited in the number of possible 
applications, but, compared with 
vocoders, the quality of articulation is 
considerably inferior. In particular, it is 
extremely difficult to produce unvoiced 
and plosive sounds. 


Modern Vocoders 

By now the reader should have gained a 
good idea of the basic principles of 
vocoding: the vocoder modulates the 
articulation of speech upon a second 
‘excitation’ signal. This is done by 
converting the input speech signal into 
data which can be used to vary the 
output signal. 

Although in principle there are various 
different ways of analysing and syn- 
thesising speech, the three vocoders 
described above are all ‘channel 
vocoders’. 

Figure 3 shows the functional block 
diagram of this type of vocoder. The 
speech signal (from the microphone) is 
fed to a bank of bandpass filters, which 
split the signal into a number of 
separate and very narrow frequency 
bands. By rectifying and feeding these 
signals through lowpass filters, a series 
of DC voltages which match the envelope 
of the filter output signals can be 
obtained. These are in fact the control 
voltages which will control the syn- 
thesiser filter bank, and represent a real 


















vocoders 


elektor april 1978 — 4-31 


Figure 3. Functional block diagram of a 
channel vocoder. All vocoders which are 
intended for musical and special effect appli- 
cations conform to this design. A real time 
spectrum analysis is made of the speech signal 
by a bank of bandpass filters and envelope 
followers. The result of the analysis is a series 
of control voltages which drive a bank of 
VCAs (voltage controlled amplifiers), to vary 
the replacement signal. Thus the spectrum of 
the original speech signal is imposed upon the 
'excitation' (normally non-speech) signal. 
The voiced/unvoiced detector continuously 
samples the speech signal and decides whether, 
at any given moment, the noise generator 
need be switched into circuit. The noise 
generator is required since most excitation 
signals do not have a sufficiently broad 
spectrum to allow the synthesis of sibilants. 
For the sake of simplicity only three channels 
are shown in the diagram. 

Photo 1. Approximately 2 years ago a 'talking' 
calculator, which contained a very small 
speech synthesiser, appeared on the market. 

Photo 2. The printed circuit board of the 
speech synthesiser inside the 'talking' calcu- 
lator. The circuit consists of only two ICs; a 
ROM which stores the speech components in 
digital form, and a micro-controller which 
selects the components for any desired word 
and, by means of a D/A converter, fits them 
together to form an analogue speech signal. 

Photo 3. An example of a mouth tube or 
mouth bag. The box contains a power ampli- 
fier and loudspeaker. The resultant sound is 
fed via the plastic tube into the mouth of the 
musician who then modulates or 'articulates' 
this signal by changing the shape of his nasal 
and mouth cavities. Thus he appears to make 
his guitar, or whatever instrument he is 
playing, speak or sing. 

Photo 4. The full-size EMS vocoder was the 
first commercially available vocoder which 
was specially designed for musical or special 
effect applications in the recording studio. 
The instrument contains several additional 
features such as a pitch extractor(pitch-voltage 
converter) and two synthesiser VCOs which 
can be played on external keyboards. 



time spectrum analysis of the speech 
signal. 

The input speech signal is also fed to a 
second circuit, the voiced/unvoiced 
detector. This continuously samples the 
speech signal to decide whether it is a 
voiced or unvoiced sound, and indicates 
the result by switching to one of two 
voltage levels (e.g. 0 V and +5 V). 

The outputs of the voiced/unvoiced 
detector and the envelope followers 
control the synthesiser section of the 
vocoder. This contains the same number 
of filters as the analyser section, so that 
the excitation signal (be it simply the 
synthesiser oscillators and noise gener- 
ator, or these two sound sources plus an 
external input) is analysed into the same 
number of separate frequency bands as 
the speech signal. Via a series of voltage 
controlled amplifiers, the outputs of the 
filter sections are then varied by the 
control voltages derived from the 
envelope followers, with the result that 
the spectrum of the speech signal is 
imposed upon the excitation signal. 

The separate channels are summed and 
fed to the output stage. The resultant 
signal possesses the ‘voice’ of the 
excitation signal (e.g. a violin), but has 
the articulation of the passage of speech. 


Furthermore, both the typical character 
of the excitation signal as well as all the 
nuances of articulation in the speech 
signal (dialect, emphasis etc.) are com- 
pletely preserved. That is to say, the 
human voice is simply replaced by that 
of whatever instrument is used for the 
excitation signal. 

In theory, therefore the voiced/unvoiced 
detector should be superfluous, however 
most excitation signals do not have a 
sufficiently wide dynamic spectrum to 
synthesise the sound of sibilants (‘s’, ‘h’, 
etc.). For this reason the voiced/un- 
voiced detector ensures that the noise 
generator provides the synthesiser 
section with the appropriate ‘raw ma- 
terial’ whenever the excitation signal 
cannot do so. 

Photos 7a and 7b show examples of 
typical signals which appear at the test 
points numbered in figure 3. The pro- 
gression of signals in photo 7a illustrates 
how the input speech signal is converted 
in the analyser section into the control 
voltages which command the VCAs. 
Photo 7b shows how the output signal is 
synthesised, using a pulse generator as 
the excitation signal. 

The second part of this article will 
contain a more detailed description of 




vocoders 


4-32 — elektor april 1978 



— S i ~ 

1 ~* ' . ■ 0 * ^ 


@ The signal after rectification. 

@ The control voltage obtained after the 
rectified signal has been ‘smoothed' by the 
lowpass filters. 

© The excitation signal from a pulse gener- 
ator. The frequency is approximately 
150 Hz. 

© This signal is obtained after the pulse 
signal has been fed through the synthesiser 
filters. 

© This is the signal which is obtained once 
the output signal of the analyser section © 
has been modulated onto the output of 
the synthesiser filters. 

® The final output signal of the vocoder is 
obtained by summing all the outputs of 
the synthesiser channels. The similarity of 
this signal with that of the original speech 
signal can be clearly seen. 


Photo 5. The Sennheiser Sound Effect 
Vocoder VSM 201. This vocoder was designed 
specifically for use in the studio, and can be 
incorporated as a module into Moog studio 
synthesisers. 


Photo 6. The 'mini' EMS vocoder; its size, 
price and extreme portability make it ideally 
suited for live stage work. 


Photo 7. These photos show the type of 
signals which typically appear at the points 
numbered in figure 3. 

© Microphone (speech) signal. The trace is 
that of the vowel 'a' in the test word 'bast'. 
© The output signal of a filter channel in the 
analyser section (centre frequency 680 Hz, 
6 dB bandwidth 140 Hz). 


how a vocoder works, and will also take 
a look at the various applications of 
vocoders. 


References: 

Figures 1, 2 and 3, photos 5, and 7: 
Sennheiser-Electronic, Wedemark, 
Hannover, West Germany. 

Photos 1 and 2: Silicon Systems Inc., 
Irvine, California 

Photo 3: Electro-Harmonix , New York 
Photos 4 and 6: EMS, London M 





formant 


elektor april 1978 — 4-33 


formant 

the elektor music synthesiser (10) l 


The COM contains a tone control ampli- 
fier with bass, middle, treble and vol- 
ume controls, and an output buffer 
capable of driving high impedance 
(> 600 f2) headphones for monitoring 
or practice purposes. The COM front 
panel also contains the indicator LEDs 
for the three power supply voltages and 
the gate signal. These indicators should 
not be regarded merely as a gimmick 
but as an important aid to monitoring 
the state of the Formant system. A fault 
in any of the supply voltages is immedi- 
ately indicated by one of the LEDs, as is 
the absence of a gate pulse. 


COM circuit 

The complete circuit of the COM is 
given in figure 1 a. 

The input signal is fed to a volume con- 
trol P 1 a and thence to an ‘anti-plop’ fil- 
ter built around Al. This is a 12 dB/ 
octave highpass filter with a break fre- 
quency of around 20 Hz. It suppresses 
low-frequency transients and rolls off 
the bass response of the system to re- 
duce ‘listener fatigue’ which can be 
caused by the low bass notes of elec- 
tronic music, especially with full bass 
boost. By rolling off the bass response 
the filter also helps protect the bass 
drivers of the loudspeakers against ex- 
cessive, very low-frequency signals. In- 
deed, if the synthesiser is to be used 
with small ‘bookshelf’ speakers it may 
be advisable to raise the turnover point 
of the filter to 40 Hz by changing the 
value of R1 and R2 to 39 k. 

The treble and bass controls, built 
around A2, are a conventional Baxandall 
network. To avoid the middle control 
interacting with the bass and treble con- 
trols it is constructed separately around 
A3. The output of A3 then feeds into a 
second volume control Plb. The use of 
a ganged volume control on a single 
signal channel may seem a little unusual, 
but it does have several advantages. A 
volume control at the input to the COM 
prevents any possibility of overloading 
Al, whatever the signal level. On the 
other hand, the provision of a volume 
control later in the circuit allows a 
better signal-to-noise ratio to be main- 
tained at low settings of the volume 


This final part of the Formant 
series completes the description of 
the synthesiser by describing the 
COM (control and output module) 
and by giving an overall wiring 
diagram. Possibilities for further 
expansion of the system are also 
discussed. 


C. Chapman 



control, since noise (principally from 
Al) is attenuated along with the signal 
as the control is turned down. The fact 
that this control produces a ‘double 
logarithmic’ characteristic does not 
cause any inconvenience in operation. 
No power amplifier is built into the 
COM as the heat generated in the out- 
put stage could cause temperature drift 
problems in other circuits in the system. 
However, the COM is provided with an 
internal output to a separate power 
amplifier, lOS. The output of the ampli- 
fier may then be brought back through 
the COM via the PA input connection 
on the COM board edge connector to a 
socket on the COM front panel (OUT 2). 
The COM output is itself also brought 
out to a socket on the front panel 
(OUT 1) into which high impedance 
headphones may be plugged. Note that 
a 6.3 mm jack socket is used for OUT 2. 
The four indicator LEDs also receive 
their power via the COM edge connector 
from the appropriate circuits, and are 
also mounted on the COM front panel. 


Construction and testing of the 
COM 

A printed circuit board and component 
layout for the COM are given in figure 2, 
a front panel design is given in figure 3 
and wiring to front panel mounted com- 
ponents is shown in figure 4. Screened 
leads should be used for the connections 
to bass, middle and treble poten- 
tiometers B, M, and T. 

Some readers may not wish to bring the 
output of a power amplifier back 
through the COM to output 2, since this 
may not be convenient especially if the 
synthesiser is to be used with, say, an 
existing hi-fi setup. In this case two op- 
tions are open. Output sockets 1 and 2 
can simply be connected in parallel or 
alternatively output socket 2 can be 
wired direct to input IS to provide an 
output signal unaffected by the tone 
and volume controls. 

It is not intended to provide a design for 
an output power amplifier since several 
good designs have already been pub- 
lished in Elektor. However, a few hints 
on the mounting of such an amplifier 
will not go amiss. As mentioned earlier, 



the power amplifier should not be 
mounted in a plug-in module since it 
may then cause thermal problems. It 
should preferably be mounted at the 
back of the module cabinet with the 
output transistors mounted on heatsinks 
whose fins are external to the module 
housing. The Formant power supply is 
not intended to supply current for a 
power amplifier, so a separate power 
supply will be required. The mains 
transformer should be mounted as far 
away as possible from the Formant 
modules to reduce hum pickup (the 
same applies to the Formant mains 
transformer). 

The COM can be tested by feeding in a 
signal from one of the VCOs and moni- 
toring it on an oscilloscope to check 
that the waveform is undistorted. The 
gain of the COM output stage, A4, can 
be varied between about 1.8 and 11 by 
means of P5. This preset should be ad- 
justed so that full drive of the head- 
phones or power amplifier is obtained 
with the volume control turned fully 
up (clockwise). 


Complete wiring diagram 

The interwiring between modules for 
the basic Formant system is given in fig- 
ure 5, but readers wishing to build a 
more extensive system can expand this 
as required. 

For clarity the supply wiring is not 
shown, but the wiring method already 
mentioned must be adhered to, i.e. each 
module should have separate supply 



Figure la. Circuit diagram of the COM, 
which consists of a tone control/headphone 
amplifier and indicator LEDs for gate pulse 
and the three power supply voltages. 

Figure 1b. Pinout of the 4136 1C. 

Figure 2. Printed circuit board and compo- 
nent layout of the COM (EPS 9729—11. 


leads from its socket back to the ‘star’ 
connection points (busbars) on the 
power supply module. The temptation 
to simplify the wiring by simply linking 
between the supply pins of the modules 
should be avoided as this will cause in- 
teraction between modules. 

The ‘Noise’ and ‘LFOs’ modules are not 
shown in figure 5, since the supply wir- 
ing is the only connection to these 
modules. 

Again for clarity, the full pinout of each 
module edge connector is not shown, 
but the connections are shown in the 
correct sequence working down from 
the top edge of each module. 

One small modification is required to 
the interface receiver p.c. board in order 
that the gate LED can be wired with 
only a single link. R30 on the interface 
receiver board is mounted in the space 
provided for D4 as shown in figure 6. 
A single wire is then connected from the 
lower pad to which R30 was originally 
connected to the appropriate pin of the 
COM socket. Without this modification 
two leads would have to be brought out 
to D4. 


Patchcords 

Due to the hardwired interconnections 
between modules, Formant is perfectly 
playable without any of the front panel 
patching sockets being used. However, 
for effects such as vibrato and tremolo, 
patchcords are used to connect the out- 
puts of the LFO module to the VCOs or 
VCA. These are very easy to make. A 







formant 


elektor april 1978 — 4-35 



Parts list for figures 1 and 2. 

Potentiometers: 

Semiconductors: 

Resistors: 

PI a, PI b • 4k7 log ganged pot. 

IC1 = 4136 (DIL package) EXAR, 

P2,P3,P4 = 100 k lin. 

Fairchild, Raytheon or 

R1.R2 = 82 k 

R3.R8.R18 = 470 n 

P5 = 220... 270 k preset. 

Texas. 

R4,R6 = 1 k5 

Capacitors: 

Miscellaneous: 

R5,R7,R11,R13 = 6k8 

Cl ,C2,C9 = 100 n 

31 -way connector to DIN 41617 

R9.R14 = 3k9 

C3,C4 = 10 n 

3.5 mm jack socket 

R10.R12 = 100 k 

C5,C6 = 39 n 

6.3 mm jack socket 

R15, R17 = 220 k 

C7 = 1 5 n 

4 collet knobs, 1 3...1 5 mm 

R16 = 22 k 

C8 = 3n3 

diameter, with pointer. 

R19 = 4k7 

C10.C1 1 ,C12 = 680 n 




4-36 — elektor april 1978 


formant 


Figure 3. Front panel layout for the COM 
(EPS 9729-2). 

Figure 4. Wiring diagram for the front panel 
mounted components. 

Figure 5. Inter-module wiring for the basic 
Formant system. Supply voltage connections 
have been omitted for reasons of clarity. The 
LFO and noise modules have been omitted as 
the only hardwired connections they have are 
supply connections. 

Figure 6. The 'gate-LED' output of the 
interface receiver can be simplified by mount- 
ing R30 in the 'D4' position. 




flexible single-core cable of about 30 cm 
length is fitted with a 3.5 mm jack plug 
at each end. The cable is soldered to the 
centre contact (ball) of the plug, no 
earth connection being necessary as the 
earth return is made through the internal 
module wiring. In the interests of long 
life the patchcord wire should not be 
too thin, and some sort of strain relief 
should be used where' the wire enters 
the plugs. About a dozen patchcords 
should prove sufficient for most appli- 
cations. Alternatively, to keep the front 
panels more tidy the patchcords can be 
made in several different lengths, each 
designed for an interconnection be- 
tween two specific modules. Different 
colours of wire may also be used to sim- 
plify checking of complicated patch 
connections. 


Front panels 

As each module was described, a suit- 
able front panel layout was also given. 
It has now been decided to make these 
panels available through the EPS printed 
circuit board service. As shown in 
figure 7, the pre-drilled metal boards are 
sprayed matt black, and the legends and 
scales are printed in white. Experiments 
have shown that this combination 












provides good legibility even under 
extreme lighting conditions. 

Further details are given in this month’s 
EPS list. 


Extending Formant 

Although the Formant system so far 
described is a versatile instrument giving 
performance comparable to commercial 
designs at a greatly reduced cost, it is 
nonetheless relatively unsophisticated 
compared to the larger commercial in- 
struments. However, because of the 
modular construction it is a simple 
matter to extend the system. Quite a lot 
can be accomplished simply by adding 
more of the modules already described, 
for example extra VCOs, VCFs and 
VCAs, to obtain a more varied sound. 
Many effects however, require the ad- 
dition of completely new modules and 
ancillary circuits, some of which it is 
hoped will be discussed in future issues 
of Elektor. One possibility which can be 
implemented immediately is the ad- 
dition of the Elektor equaliser (January 
1978) to allow presettable tailoring of 
the synthesiser spectrum. The equaliser 
p.c. board is of Eurocard format, com- 
patible with the other Formant mod- 
ules. Another module which it is hoped 
to feature, which will greatly increase 

















4-38 — elektor april 1978 


formant 



the tone colour possibilities of the sys- 
tem, is a 24 dB/octave VCF module. 
Banks of resonant filters are also a use- 
ful addition to the tone-forming capa- 
bilities of the synthesiser, especially for 
the production of vocal-type sounds. 
These are not voltage-controlled filters, 
but have manually presettable centre 
frequency and Q factor. 

Phasing circuits are frequently used in 
synthesisers, and are particularly useful 
for more realistic reproduction of string 
tones. Another tone modifying circuit 
which is often used is the ring modu- 
lator. This circuit produces the sum and 
difference of two input frequencies at 
its output. The frequencies produced 
are not necessarily harmonically related, 
and the sound is not particularly 
‘musical’; however, the ring modulator 
is extremely useful for special effects 
such as bells, gongs and cymbals. 

In its basic form the range of expression 
available from the synthesiser is some- 
what limited by the fact that it is played 
by a keyboard. However, there are 
various ways in which this can be 
remedied. The addition of a ‘pitch- 
bender’ joystick, which feeds a manu- 
ally controllable DC voltage to the 
VCOs, allows modulation of the pitch 
of a note by hand in much the same 
way that a guitarist "pulls’ the strings of 
his guitar. 

An interesting possibility is the elimin- 
ation of the keyboard by playing the 
synthesiser via another instrument. This 


is accomplished by the use of a pitch-to- 
voltage converter, which produces an 
output voltage proportional to the pitch 
of the control instrument. This in turn 
controls the frequency of the syn- 
thesiser VCOs. An envelope follower 
produces an output voltage which fol- 
lows the control instrument’s amplitude, 
and this is used to control the gain of 
the VCAs. The result is a synthesiser 
sound which has the dynamics of the 
original instrument. 

Other useful additions to the synthesiser 
system are sequencers, sample-and-hold 
circuits and reverberation/echo units. 
Sequencers are used to store (either by 
analogue or digital means) a sequence of 
VCO/VCF control voltages. These are 
then ‘played back’ into the synthesiser 
automatically to generate a note se- 
quence which can, for example, be used 
to provide the backing for a manually 
played melody. 

A sample-and-hold circuit is frequently 
used to take sequential samples of the 
instantaneous voltage of a sawtooth 
waveform. This sequence of voltage 
samples is used to control the syn- 
thesiser to generate a pseudo-random 
sequence of notes. 

Reverberation units are used to enhance 
the somewhat ‘dry’ sound of the syn- 
thesiser by allowing notes to die away 
gradually rather than be cut off ab- 
ruptly when a different key is pressed. 
Such units may contain mechanical de- 
lays such as plates or springs, or purely 


Figure 7. The complete set of front panels is 
now also available through the EPS service. 


electronic delays such as analogue shift 
registers may be employed. 


Loudspeakers 

Before concluding this article a few 
words on the choice of loudspeakers for 
use with Formant will not come amiss. 
Readers building a synthesiser for home 
use will probably wish to play the in- 
strument through an existing hi-fi setup, 
at least to begin with. If this is the case 
care should be taken not to overload the 
loudspeakers, by keeping the volume 
fairly low. Hi-fi loudspeakers are de- 
signed to handle a much more broadly 
distributed power spectrum than that 
produced by a synthesiser, and it is 
quite easy to damage the tweeters with 
a sustained high frequency note. 

For serious use a purpose-designed loud- 
speaker system should certainly be con- 
sidered. Horn systems are to be favoured 
because of their high efficency and a 
dealer who specialises in electronic 
music systems should be able to offer 
advice on suitable loudspeakers. M 


loudspeaker connections 


elektor april 1978 — 4-39 



Many hi-fi enthusiasts may not realise 
that significant distortion may be intro- 
duced into an audio signal by the con- 
nections between the amplifier output 
and the loudspeakers. In the first place, 
output current from the amplifier has to 
travel across several non-soldered metal- 
to-metal contacts, for example plug and 
socket connections at the amplifier out- 
puts and the loudspeaker inputs, and 
loudspeaker switches within the ampli- 
fier (of which more later). For minimum 
distortion these contacts should not 
only have a very low resistance, but 
must also have a constant, linear resist- 
ance. 

Oxidation of the metal surfaces of 
plugs, sockets and switch contacts can 
produce a non-linear resistance which 
varies with the current flowing through 
it, thus distorting the signal fed to the 
loudspeakers. DIN loudspeaker plugs 
and sockets are particularly bad in this 


respect due to their very small contact 
area, and should be avoided. Where non- 
soldered connections must be made the 
use of screw terminals or robust 4 mm 
‘banana’ plugs and sockets is to be 
preferred. 

The second area which can cause degra- 
dation of the audio signal is the con- 
necting cable itself. When a loudspeaker 
is being driven by an amplifier the loud- 
speaker cone should move exactly in 
sympathy with variations of the ampli- 
fier output voltage. Ideally, if a loud- 
speaker is fed with, say, a step input, 
the cone should move quickly to the 
appropriate position and stop. In prac- 
tice, of course, this does not happen. 
A loudspeaker possesses inertia and 
compliance, so that the cone will tend 
to oscillate about its final position 
before settling down. Whilst this 
‘ringing’ is in progress the loudspeaker 
acts as a generator and tries to pump 
current back into the amplifier output. 
If the amplifier output impedance is 
low (and it generally is) the loudspeaker 
sees a short-circuit and the cone move- 
ment is quickly damped by electro- 
magnetic braking. The ‘damping factor’ 
of an amplifier is defined as the ratio of 
the load impedance to amplifier output 
impedance. As the output impedance 
of a modern transistor amplifier is 
generally a fraction of an ohm, damping 
factors are typically between 50 and 
200 with an 8 ohm load. However, the 
resistance of the loudspeaker connecting 
cable appears in series with the amplifier 
output and must be considered as part 
of the amplifier output impedance. If 
the loudspeaker cable is thin its resist- 
ance will be high and the damping 


factor will be considerably reduced. In 
addition, some of the amplifier’s output 
voltage will be dropped across the cable 
resistance rather than appearing across 
the loudspeaker. 

Thus the second rule when connecting 
loudspeakers is to use heavy-duty cable. 
Fuses, which are sometimes inserted in 
series with amplifier outputs for loud- 
speaker protection, should also be 
avoided since they can have a significant 
resistance. 

Recent research, particularly by Japanese 
manufacturers, seems to indicate that 
the inductance of loudspeaker cables 
has a significant effect on transient 
response, and Hitachi, JVC, Pioneer and 
Sony are all introducing special loud- 
speaker cables which are claimed to 
give an improved sound. Whether or not 
these claims are true is still a matter for 
conjecture. 

Returning to the subject of loudspeaker 
switching, figures 1 and 2 show two 
typical switching arrangements which 
allow two sets of speakers to be con- 
nected to an amplifier, either indepen- 
dently or simultaneously. One channel 
only is shown and the circuits are ident- 
ical for the other channel. Although 
such switching arrangements offer con- 
venience of use, they may not be such a 
good idea from a sound quality point of 
view due to the contact resistance of the 
switches. If loudspeaker switching is 
employed in an amplifier then the 
switches used should be rated at several 
amps to ehsure minimum contact resist- 
ance. 

Both the switching arrangements shown 
in figures 1 and 2 have their advantages 
and disadvantages. In figure 1 >oth 
speakers appear in parallel across the 
amplifier output in the A + B position . 
Whilst this does mean that the camping 
factor is maintained the reduced load 
impedance can cause overloading. 

In figure 2 the speakers are connected 
in series in the A + B position. Assuming 
that both speakers have the same 
impedance this connection, of course, 
doubles the load impedance, so there 
is no risk of overload. However the 
available output power is halved (since 
P = U 2 /R) and the damping factor is 
reduced to less than unity, since each 
loudspeaker has the other in series with 
it as a source impedance. 

In conclusion, anyone contemplating 
the building of an audio amplifier and/ 
or loudspeakers would be well advised 
to bear in mind all the points raised in 
this article. To summarise: 

1 . Connection to the loudspeakers 
should be made with the minimum 
number of non-soldered connections 
(plug and socket connections and 
switches) in series with the signal path. 

2. The cable to the loudspeakers should 

have as low a resistance as possible. 
Fuses in series with the loudspeakers, 
although seemingly desirable from a cir- 
cuit protection point of view, have a 
detrimental effect on sound quality and 
should be avoided. K 


s 1a s 1b 



S ja S 1b 

\ 

\ \ 






4-40 — elektor april 1978 


market 



16 segments 

The entire alphabet from A to Z, 
all digits from 0 to 9, plus, minus, 
equals and summation signs and a 
whole series of other symbols 
may be depicted by a 16-scgment 
display, which Siemens is now' 
putting on the market under the 
designation HA 4041. This LED 
display offers an alphanumeric 
set comprising 64 characters each 
four millimeters high. Four such 
displays are combined on one 
module with the associated 
electronics. The modules can be 
arranged in rows of practically 
unlimited length. 

The so-called 7-segmcnt displays 
with three horizontal and four 
slightly sloping vertical bars arc 
well-established in a wide field of 
applications ranging from 
measuring instruments to TV sets 
and watches and clocks of all 
sizes. Liquid crystals and oblong 
light-emitting diodes are equally 
suitable as the display medium. 
However conventional 7-segrnent 
displays essentially provide for 
the ten digits only. 



The new HA 4041 module 
incorporates extensive electronics 
for four 16-segment displays, 
which should markedly increase 
the presentable information. Each 
module contains a decoder for the 
64-character ASCII set, a 
multiplexer, a memory and the 
LED driver stages. Externally, the 
circuits behave like a RAM device. 
The operating voltage of 5 V 
permits easy interfacing with, for 
example, TTL. 


A very significant feature is 
compatibility with micro- 
processors, which could now learn 
to talk with the aid of the 
16-segment displays - an 
economically unjustifiable 
proposition so far. The 
alphanumeric character set 
permits the display of operating 
states and program progress. A 
related application would be in 
keyboard stations or phototype- 
setting equipment, w'here the 
typed information could be 
displayed for checking purposes 
before it is printed. 

Each of the four display units of a 
module puts out 0.1 med per 
segment with a viewing angle of 
20° from all sides. Up to 
16 displays (four HA 4041 
modules each with a width of 
25 mm) may be lined up as an 
array, while more than 16 displays 
can be combined with modest 
circuit requirements. The module 
packages are designed to present 
an evenly spaced display, also 
when used in arrays. 

Siemens AG, Postfach 103, 
D-8000 Miinchen 1, 

Federal Republic of Germany 

(696 M) 


Dot matrix printer 

A range of compact dot matrix 
impact printers, the 7040 series, 
has been introduced by 
Impectron Limited. Suitable for a 
wide range of data output 
applications such as point of sale 
documentation and data logging, 
the low cost range features ease of 
operation, compact size and 
simple interface circuitry. 

The printer utilises a serially 
driven print element consisting of 
7 print solenoids and associated 
print wires. The wires are arranged 
in a vertical line and driven 
horizontally across the paper at 
constant speed. Print speed is over 
one complete 3.3” line per second, 
and each line contains 
40 characters. Character height is 
0.123 inches. 

By use of external control 
circuitry, the printer may produce 
characters of almost any density 


or fount desired. Because the 
print head travels at constant 
speed, there is no need for a 
complex feedback system to 
determine the correct timing of 
print pulses. Ribbon feed, ribbon 
reverse (and in some cases paper 
feed) are controlled automatically 
without control signals. 

Unlike many printers, the 
7040 series has no clutches, 
timing discs or reversing 
mechanisms to control print head 
movement. The head always 
traverses a complete line from left 
to right and then returns to its 
home position regardless of the 
number of characters printed on 
each line. 



The range has been designed to 
produce up to 5 copies of the top 
copy, depending on paper 
thickness and type. Maximum 
paper thickness is 0.015” overall. 
Two models are currently 
available from Impectron. The 
basic model 7040 is a simple 
printer with no paper supply or 
document handling mechanisms. 
Optional extras do however 
include paper roll holder, journal 
take up and assembly, or 
secondary motor for high speed 
paper feed. 

A more sophisticated variation, 
the Model 7040T, is arranged in a 
flat bed document printer 
configuration. As documents are 
inserted for printing, a solenoid- 
activated roll clamps the 
document, permitting proper 
feeding and preventing accidental 
or premature removal. This model 
also features a highspeed 
document feed of 10 lines per 
second. Optional features of the 
7040T include reverse document 
feed and top of form sensors. 

Specification: 

Speed: 

Character Height: 

Data Input: 

Line Length: 

Print Solenoid 
Power: 


1 .04 lines/ 
second 

0.1 23 inches 
Synchronous 
3.3 inches 

40 V (DC) 

± 10% @3.6A 
peak current 
per solenoid. 
Average cur- 
rent approx. 
0.87A for 

1.5 ms cycle 
time. 


Impectron Limited, 

Impectron House, 

23-31 King Street, 

London W3 9LH, England 

(716 M) 


Remote control chips 

AMI Microsystems have 
introduced a 31 -command remote 
control chip set with keyboard 
inputs, oscillators, and both 
analog and digital receiver output 
all on board the chip. 

Consisting of an S2600 transmitter 
and an S2601 receiver, the set 
reduces the part count in 
equipment designed for remote 
control via radio frequency, 
infrared, ultrasonic or hardwire 
transmission media. Among the 
applications for the devices are 
motorized toys such as trains and 
boats, home security systems, 
automatic telephone calling 
equipment, industrial controls, 

TV and stereo controls, and 
traffic controls for emergency 
vehicles. 

The AMI S2600 and S2601 have 
eliminated the need for external 
crystals; only a resistor and a 
capacitor are required externally 
for a frequency reference. The 
S2601 receiver will tolerate up to 
± 24% difference in the timing 
frequency and still operate. 
However, the circuit has a very 
high immunity to noise or 
spurious commands. 

Spurious command rejection has 
been achieved through a 5-bit 
command code system w'hich 
requires that identical, proper 
commands be transmitted twice 
in succession before the receiver 
issues an output. In addition, a 
correct five-bit fixed (mask- 
programmable) preamble code 
must be received. 



Eleven outputs (six digital, three 
analog, a pulse train and an on/off) 
arc available from the receiver. 
Five binary outputs present the 
five-bit command code received; 
the sixth digital output is a ‘data 
valid’ signal. The pulse train is 
useful for indexing a stepping 
switch, as in TV channel selection, 
or operating a stepping or 
counting device in industrial 
controls or toys. The on/off 
output can be used to remove and 
restore the main power supply. 
The analog outputs can 
independently provide up to 
64 distinct DC levels for 
controlling motor speed, volume, 
brightness, or similar electronic 
settings; one of these analog 
outputs is mutable and can be 
used for TV sound control. 





market 


elektor april 1978 - E 15 



The S2600 transmitter is a low- 
power drain CMOS chip 
(dissipating only 20 mW) with an 
on-chip oscillator, 1 1 keyboard 
inputs, a keyboard encoder, a 
shift register and control logic. Its 
output is a 40 kHz square wave 
which is pulse code modulated. 
The S2600 can transmit a 12-bit 
message including sync frame, 
preamble, 5-bit command code, 
and end of message bits every 
38.4 milliseconds. 

The S2601 receiver is a P-channel 
MOS chip with on-chip oscillator, 
five keyboard inputs, a 40 kHz 
• ignal input, decoding logic and 
eleven outputs. Its on-chip 
memory saves received commands 
md the logic compares them with 
later receptions. If the codes do 
not match, the receiver saves the 
iast code received for its next 
comparison try. When two 
successive identical codes are 
received, a valid output is issued. 
AMI Microsystems Ltd., 

!08A Commercial Road, 
Swindon, Wiltshire, England. 

(718 M) 


Digital inductance meter 

AIM Cambridge Ltd is pleased to 
announce a new, low cost 
Automatic Digital Inductance 
Meter type DLM307. This 
instrument, intended as a 
complement to the Digital 
Capacitance Meter DCM302, is 
rally auto-ranging and will 
measure up to 1.999 H full scale, 
i hilst the most sensitive range has 
i resolution of 1 pH. The readout 
j a 3’A digit LED display and the 
anit can be powered either from 
catteries or a small mains power 
anit, both of which are supplied. 
\s in AIM’s Digital Capacitance 
Meter, the unit has no operator 
controls apart from a touch-pad 
vhich is used to turn on the 
nstrument for about 10 seconds. 
The correct measurement range is 
round automatically within 
1 seconds of switch-on, after 
vhich measurements are repeated 
rrery 0.4 seconds. 

The measurement technique 



employs a highly linear L-R 
oscillator whose period is 
proportional to the value of the 
inductor being measured. This 
period is then measured using 
digital techniques employing 
CMOS integrated circuits. One 
advantage of this approach is 
that the measurement frequency 
is appropriate for the value of 
inductor being measured. For 
example, large value inductors 
are measured at frequencies 
between 25 Hz and 250 Hz whilst 
small inductors are measured at 
frequencies up to 1 MHz. 

Aim Cambridge Limited, 

Edison Road, Industrial Estate, 

St. Ives, Huntingdon, Cambs. 

PEI 7 4LF England. 

(715 Ml 


Microwave spectrum 
analyser 

The new 7L18 microwave 
spectrum analyser from Tektronix 
incorporates several advanced 
technological innovations to offer 
a combination of exceptional 
performance and ease of 
operation. A high-stability phase- 
lock system yields a resolution of 
30 Hz at frequencies up to 
12.5 GHz, while external 
waveguide mixers extend the 
overall frequency range up to 
60 GHz. Other important 
technological developments used 


in the 7L18 include micro- 
processor-aided controls for ease 
of operation and adjustment, a 
split digital-storage system, and 
YIG tuned filters for spurious-free 
display from 1.5 GHz to 18 GHz. 
The 7L18 is a three-module wide 
plug-in unit for the Tektronix 
7000 Series modular 
instrumentation range. With a 
direct coaxial input it will display 
the spectrum of signals from 1.5 
to 18 GHz, with a resolution of 
30 Hz up to 12.5 GHz. 

The new external waveguide 
mixers extend the frequency 
coverage to 60 GHz with a 
response flatness specified at 
± 3 dB or better. Hence relative 
amplitude measurements can be 
made with confidence w hen 
operating with waveguide mixers. 
In addition, the built-in 
preselector system is fully 
characterised for absolute 
amplitude measurements up to 
18 GHz. 

Measured in terms of residual 
frequency modulation, the 
stability resulting from the new 
phase-lock circuitry is specified as 
10 Hz or less up to 4.5 GHz 
(about four parts in 10 8 ). 

Digital storage provides flicker- 
free displays at the lowest sweep 
speeds, fine detail and unlimited 
storage time for subsequent 
viewing, comparison or easy 
photographic recording. A split 
memory allows comparison of a 


reference with an existing 
spectrum or a calculated display 
of the difference between two 
spectra. The storage circuitry also 
includes a maximum-hold 
capability that allow s monitoring 
of frequency or amplitude signal 
variations. 

A microprocessor provides 
automatic resolution and sweep 
time/division modes to optimise 
setting up the display and prevent 
many potential operator errors. In 
the non-auto mode of operation, 
any combination of control 
settings which results in an 
uncalibratcd display also turns on 
the 'uncalibratcd’ indicator light. 
The 7L18 spectrum analyser is 
easily transportable, and 
applications include microwave- 
relay and satellite communication, 
frequency management and 
microwave component and 
system manufacture. 

The instrument can also be 
converted to a high-quality 
microwave receiver for time- 
domain measurements by setting 
the frequency span to zero and 
using the calibrated time base. 
Tektronix U.K. Ltd., 

Beaverton House, P O Box 69, 
Harpenden, Herts, England. 

(717 Ml 






E 16 — elektor april 1978 


market 



P.C.B. KIT 

Mega Electronics Ltd, introduced 
a comprehensive kit which 
enables the preparation of 
artwork for, and the production 
of, both printed circuit boards 
and boards and front panels or 
labels. 



Known as the Photolab Kit, it 
consists of an ultraviolet exposure 
unit, drafting aids and film, 
positive resist coated epoxy glass 
laminate sheets, developing and 
etching trays, label and panel 
materials, high-speed drill, and all 
the requisite developers. 

The Photolab Kit has been 
designed for use by both the 
hobbyist and the professional 
engineer. It has been introduced 
to fill the gap between commonly 
used T-off prototype p.c.b. 
production methods and the 
facilities offered by the existing, 
larger kits currently available. It is 
priced at only £ 44.50, complete, 
and can handle p.c. boards and 
labels of up to 9 x 6 in. 

Mega Electronics Ltd, 

9, Rad win ter Road, 

Saffron Walden, Essex CB11 3HU, 
England 

(682 M) 


ptP based analyser 

The 7L5 from Tektronix U.K. 

Ltd. is a microprocessor-based 
spectrum analyser that achieves 
exceptional frequency accuracy 
(two parts in 1 0 6 ) through a 
unique combination of synthesiser 
and digital technology. The 
inherent stability of the 
synthesiser method used, coupled 
with digital tuning techniques, 
means that the centre frequency 
can be set with 6-digit accuracy 


immediately after turn-on, with 
no need to fine-tune the displayed 
signal. 

The built-in microprocessor 
intelligence is used to simplify 
operation of the instrument. 
Internal processing decodes 
control settings, processes 
frequency and reference level 
information and optimises sweep 
time and resolution for the 
chosen frequency span. At turn- 
on, the 7L5 is preset to a 
reference level of +17 dBm and a 
centre frequency of zero, which 
provides not only input 
attenuation to protect the front 
end but also a marker to verify 
correct operation. 

The 7L5 spectrum analyser has a 
full 80 dB spurious-free dynamic- 
range for measuring wide relative 
amplitudes. Nanovolt sensitivity 
allows very low-level signals and 
noise to be measured. A front- 
panel input buffer control greatly 
increases front-end immunity to 
intermodulation, while 
maintaining a constant reference 
level. 



The instrument is fully calibrated 
in dBm, dBV or volts per division. 
The reference level can be set in 
1 dB steps, eliminating the need 
to interpolate amplitude levels. 

To accommodate a wide variety 
of input impedances, the 7L5 uses 
plug-in modules. Three standard 
modules offer 50 ft, 75 ft and 
1 MB, and special modules for 
other impedances can be provided. 
The 7L5 incorporates a digital 


storage and averaging technique. 
The entire display is stored 
electronically and updated during 
each sweep, and two complete 
displays can be held in the 
memory for comparison. Two 
display modes are available: a 
conventional peak display or a 
digitally averaged display. For 
special measurements, such as 
signal/noise ratio, these two 
modes can be used simultaneously 
by setting the continuously 
adjustable peak/average threshold. 
A 'maximum hold’ control 
enables maximum signal levels to 
be stored for checking long-term 
amplitude and frequency drifts. 
Options available for the 7L5 
spectrum analyser include a 
tracking generator, logarithmic 
frequency display and front-panel 
readout unit. 

Tektronix U.K. Ltd., 

Beaverton House, P O Box 69, 
Harpenden, Herts, England 

(678 Ml 


Chiming annunciator 

A new chiming audible has been 
added to the Highland range of 
audible warning devices. 

The repeating chime has a 
frequency of 2900 Hz ± 500 Hz 
working at 6 . . . 16 V DC. 



Maximum current used at 16 V 
DC is 8 mA. Light in weight and 
very compact it is easily installed 
with a single 32 mm hole fixing. 
Overall dimensions are 42.9 mm 
(1 11/16") diameter at back, 

47.6 mm (1 7/8") in length. It is 
simply connected by two screw 
cable connections. 

Chiming Sonalerts produce a 
unique tone by electronic means. 
A semi-conductor oscillator 
driving a piezo-ceramic transducer 
produces a penetrating but not 
unpleasant sound. The audible 
volume can be varied by adjusting 
the supply voltage. 

Highland Electronics Ltd., 
Highland House, 8 Old Steine, 
Brighton, East Sussex, 

BN1 I EJ England 

(683 Ml 


5 Amp negative voltage 
regulator 

This hybrid voltage regulator, the 
pA79HG, is an adjustable four 


terminal device capable of 
supplying in excess of 5 A over a 
24 V to -2.2 V output range. 
It is a complement to Fairchild’s 
pA78HG positive adjustable 
regulator. 



It has been designed with all the 
inherent characteristics of the 
monolithic four terminal regulator 
and offers full thermal overload 
plus short-circuit protection. 
Should the safe operating area 
ever be exceeded, the device 
simply shuts down. 

Packaging is in a hermetically 
sealed TO-3 can. Absolute 
maximum ratings include an input 
voltage of -40 V and internal 
power dissipation of 50 W at a 
case temperature of 25° C. 
Fairchild Camera & Instrument 
(UK) Ltd., 230 High Street, 
Potters Bar. Herts, EN6 5BU, 
England 


(681 M) 


Elektor announces 
new logic 

Beek, Limburg, 1 Apr. 1984- 
Elektor, one of the world’s major 
suppliers of kits and components, 
today announces a new line of 
feminine logic which is now- 
available from stock to Elektor 
readers. 

The first device in this new logic 
family is called the 'Maybe’ gate. 
Its new logic symbol is shown 
here. 



The device functions as follows: 

(a) inputs 1 and/or 2 ‘high’, may- 
cause the output to go ‘high’ 
(but maybe not) 

(b) if the output does go ’high’ it 
will remain ‘high’ unless it 
goes ‘low’ 

(c) If the output is ‘high’ and 
either input 1 or 2 goes ‘low’ 
the device will probably go 
Tow’ 

Elektor is certain you can see the 
potential for the ‘Maybe’ gate in 
such items as 'household 
computers’, ‘computer-piloted 
automobiles’ and, most important 
of all, w eather-forecasting com- 
puters. 

Elektor . Peter Treckpoelstraat 2, 
6191 VK Beek (L), Netherlands. 

(695 Ml 





I 



Wherever possible in Elektor circuits, transis- 
tors and diodes are simply marked 'TUP' 
(Transistor, Universal PNPl.'TUN' (Transistor, 
Universal NPN), 'DUG' (Diode, Universal Ger- 
manium) or 'DUS' (Diode, Universal Silicon). 
This indicates that a large group of similar 
devices can be used, provided they meet the 
minimum specifications listed in tables la and 
1b. 


Table 6. Various equivalents for the BC107, 
-108, . . . famines. The data are those given by 
the Pro-Electron standard; individual manu- 
facturers will sometimes give better specifi- 
cations for their own products. 


Remarks 



BC 107 

BC 208 

BC 384 

BC 108 

BC 209 

BC 407 

BC 109 

BC 237 

BC 408 

BC 147 

BC 238 

BC 409 

BC 148 

BC 239 

BC 413 

BC 149 

BC 317 

BC 414 

BC 171 

BC 318 

BC 547 

BC 172 

BC 319 

BC 548 

BC 173 

BC 347 

BC 549 

BC 182 

BC 348 

BC 582 

BC 183 

BC 349 

BC 583 

BC 184 

BC 382 

BC 584 

BC 207 

BC 383 



Table 5. Minimum specifications for the 
BC107, -108, 109 and BC177, -178. -179 

families (according to the Pro-Electron 
standard). Note that the BC179 does not 
necessarily meet the TUP specification 
He, max = 50 mAI. 


169/259 

^max = 

50 mA 


251 .. . 253 
low noise 


'cmax = 
200 mA 


Table 3. Various transistor types that meet the 
TUP specifications. 


TUP 

BC 157 

BC 253 

BC 352 

BC 158 

BC 261 

BC 415 

BC 177 

BC 262 

BC 416 

BC 178 

BC 263 

BC 417 

BC 204 

BC 307 

BC 418 

BC 205 

BC 308 

BC 419 

BC 206 

BC 309 

BC 512 

BC 212 

BC 320 

BC 513 

BC 213 

BC 321 

BC 514 

BC 214 

BC 322 

BC 557 

BC 251 

BC 350 

BC 558 

BC 252 

BC 351 

BC 559 


BC 107 
BC 108 
BC 109 


BC 177 
BC 178 
BC 179 


The letters after the type number 
denote the current gain 
A a' Ij3, h fe ) = 125-260 

3 a' = 240-500 

: a' = 450-900. 


100 mA 
1 00 mA 
100 mA 

300 mW 
300 mW 
300 mW 

150 MHz 
150 MHz 
150 MHz 


100 mA 
1 00 mA 
50 mA 

300 mW 
300 mW 
300 mW 

130 MHz 
130 MHz 
130 MHz 











































































E 18 — elektor april 1978 









giPPiglglpi 

■EMIT 

lisISsm 




gjgjlgp|P!BBW| 



iDiBiBiqiQiDiDi 



1 | j uoio 

Pin-compatible CMOS equivalents available from Teledyne Semiconductor and National Semiconductor 






















4011 4012 





































E 20 — elektor april 1978 


advertisement 



Electronic Components Specialists 

325, DUTOIT STREET, P.O. BOX 2749, PRETORIA, 

0001 TRA NSVAAL. REPUBLIC OF SOUTH AFRICA 
I Telegraph address "TRINITRON" 


Stockists of 


Heathkit — Siemens — 
Monsanto — 
Texas — Interface — 
Linear — MOS — Signetics 
Fairchild 
PRODUCTS 


Crystals & ^ 
Sockets 
Piezo Electric 
Ceramics 


ELEKTOR 

MAGAZINES 

BINDERS 

PRINTED CIRCUIT 
BOARDS 


Transistors 

Diodes 


Capacitors 

Tantalum 

Polyester £ 

Electrolytic — 

Co-axial & 

P.C. Mount - 


I.C.'S Op Amps 

a Vol. Regs., 

Lin. Amps, 
Digitals, Nor, 
Latching etc. 
etc. 


Electronic books, 
Projects, data etc 


Panel Meters 


Toggle switches 
Connectors . . . 
Potentiometers 
Trim pots . . . 


Inductors 


SCRS . . . Triacs . 
Quaracs . . . Variacs 
Transformers 


Sole South African distributors for ELEKTOR 


Trade Enquiries Welcome